Michael Hornsby, Impact Manager at Anti-Corruption Data Collective (ACDC) and David Szakonyi, ACDC co-founder, discuss the value that cross-sectoral collaborations might bring to complex, cross-border investigations into illicit financial flows and kleptocracy using leaked, commercial and public data sets. 

Picture the scene: you’re an investigative journalist and, out of the blue, a source offers you a treasure trove of data that exposes the inner workings of an opaque financial jurisdiction long suspected of enabling large-scale financial crime and corruption.  

Who would you call? 

For many journalists in this position, the answer has been an international media organisation with the ability to set up and coordinate collaboration across multiple newsrooms in multiple countries. This model has successfully underpinned many of the largest and most transformative cross-border collaborative investigations into offshore finance and corruption, such as the Panama Papers, Troika Laundromat and Pandora Papers. 

The involvement of diverse journalistic expertise from around the world helps to ensure that the most interesting and timely stories are extracted from the data. These massive collaborative projects have led to equally enormous impacts: toppling governments, sparking prosecutions, informing new sanctions, regulations, legislation and international frameworks, as well as catapulting the offshore world into the public consciousness.  

Despite this arguably successful track record, some journalists, like Frederik Obermaier, who together with Bastian Obermayer, received the Panama Papers and Pandora Papers (along with other high-profile leaks) have wondered how non-journalistic actors in the anti-corruption space, such as academics and civil society experts, can contribute to even greater impact by getting involved in collaborative investigations early on. 

New models for collaboration  

In November, the organisations behind two current Governance & Integrity Anti-Corruption Evidence (GI ACE) projects, along with the National Endowment for Democracy, convened experts from across journalism, technology, policy and academia at a Data against Kleptocracy workshop held in London. Our discussion focused on how journalists and data-savvy social scientists can better collaborate to investigate illicit financial flows and kleptocracy using leaked, commercial and public data sets.    

For one of these GI ACE projects, our organisation, the Anti-Corruption Data Collective (ACDC), is collaborating with the Platform to Protect Whistleblowers in Africa (PPLAAF) to mine the 2021 Congo Hold-Up leak from BGFI Bank in the DRC for fresh insights. Journalists from 18 countries pored over this dataset – the largest ever leak of financial records from the African continent – and exposed how the DRC’s former President Joseph Kabila and his inner circle laundered millions of dollars through the bank. 

The initial results from this collaboration are helping to prove the concept that we set out to test when Frederik Obermaier and David founded ACDC in 2020, together with the anti-corruption policy expert Zoë Reiter. By applying the kinds of big data analysis that informs much of contemporary social science research, we have been able to generate insights which were not available during the original wave of Congo Hold-Up reporting. For instance, by identifying previously unrecognised companies as highly likely to have been involved in money laundering through BGFI’s branch in the DRC. 

Data science approaches further allow us to evaluate the effectiveness of international banks’ de-risking policies in reducing money laundering and the banks’ exposure to it. We are also able to determine whether whistleblower revelations changed attitudes and appetites toward risk at BGFI in the DRC.  

Reaching this stage of the collaboration has meant overcoming several challenges, many of which were familiar to participants in the London workshop: Journalists and academics work on different timelines, they ask different questions of the data in front of them, and their organisations have different understandings of the ethical and legal questions around public interest, privacy and data processing.  

However, this nascent mode of collaboration has already seen some stand-out successes.  

What can we learn from early success stories? 

Matthew Collin, an economist at the EU Tax Observatory who is part of our team for the Congo Hold-Up project, has analyzed publicly available data from the Cayman National Bank and Trust (Isle of Man) that was initially received by the journalist collective Distributed Denial of Secrets. His work produced three new discoveries, helping pull back the curtain on the potential level of wealth contained in offshore accounts, the frequency of politically exposed persons (PEPs) controlling these accounts, and the critical importance of shell companies in providing the opacity to make it all happen.  

Our colleagues in the current GI ACE cohort from University of Exeter, OCCRP and University of Oxford have launched a similarly ambitious project that taps over 50 terabytes of administrative and leaked datasets. By developing new tools to process such massive reams of information, they’re contributing not just to our academic understanding of the role of professional enablers in facilitating illicit financial flows around the world, but also creating real value for journalists working with similar data down the line. 

The discussion in London made it clear that individuals, organisations and institutions in this space need varied types of support in order to pursue more collaborations, whether they receive and archive leaks, coordinate investigations or conduct academic research. Many teams are facing the same thorny problems in analyzing these large datasets, and there was widespread interest in creating platforms and workshops to better communicate lessons learned.  

But there is also a mismatch between the wealth of data waiting to be analyzed and the resources available to teams to pilot potential ideas and make their cases to journalists. GI ACE is an exemplary example of a program providing much-needed support to projects such as ours so that we can demonstrate the importance of these types of cross-sector collaborations. 

Additionally, to achieve the added value promised by academic insights, it is  crucial to bring empirical findings to policy makers and practitioners to help underpin and drive effective action. This often falls outside of the normal practice of both academics and journalists, highlighting the important role of civil society organisations and impact-oriented partnerships in making the bridge to end users.  

This nascent community of interdisciplinary practitioners will continue learning from each other and we see a clear mandate for developing this collaborative work further. Our hope is that in the future, one of the first calls a journalist will make after receiving a leak might be to a university. 

Originally posted as a Governance & Integrity Anti-Corruption Evidence (GI ACE) Programme blog.