By Dr. David Soud
Over the past several years, I.R. Consilium has earned a strong global reputation for analyzing, mapping, and targeting what we have called “invisible supply chains” – the routes by which bad actors move normally legitimate goods that have been turned to criminal ends. As we have revealed in multiple reports, such supply chains are often hidden in plain sight. Careful sifting and correlating of trade data, open source information, and interviews with well-placed sources can at least begin to reveal what is being moved where, in what form, for what purpose, by whom.
This is painstaking work. It involves drilling down into reams of data at different levels of aggregation. You have to account for potential errors and blank spots in those data. You have to make sure that what you find in one information stream matches up with, or at least doesn’t contradict, what turns up in another. You have to remember that a signal that “pops” amid all that information can signify many things, most of them benign. You have to know precisely where insight ends and apophenia begins, and how the worst move you can make is to buy into an appealing narrative and try to make the data confirm it. And when you do make a mistake, even a seemingly insignificant one, you have to be willing to own it and correct it as quickly and clearly as possible.
But you also need something else, which is too often ignored by analysts and journalists who have an unexamined faith in data and a desire to make something happen. That something else is subject matter expertise. Without it, you can’t even know the right questions to ask.
If you’re going to analyze the behavior of a cargo vessel, you had better know something about the complexities of the maritime space, and the many reasons why a ship might appear to loiter, or even to turn off its Automatic Identification System (AIS) transponder. If you’re going to comment on oil theft in Nigeria, it would behoove you to know something about crude oil and Nigerian crude in particular, and about the complexities of power, privilege, and patronage in that country. And if you’re going to report on gold smuggling out of Sudan, you ought to have done your homework on how the gold supply chain works, not just in terms of numbers but through actual physical processes.
That last topic provides a useful lesson. In July 2022, CNN released a report titled “Sudan’s Gold, Russia’s War.” The report investigated the role of the Wagner Group in illicit flows of gold out of Sudan, arguing that Sudanese gold was helping finance Russia’s invasion of Ukraine. At its core, the story was a good one: there is no question that Wagner is involved in those flows as well as ones in other African countries, and that its involvement in Sudanese gold connects with Russian support for the regime there. The report included impressive video of a visit to the gold processing plant at Al-Ibaidiya, and compelling documentary evidence of the facility’s still being beneficially owned by Russian interests linked with Wagner.
Russia, one of the world’s top three gold producers with extensive central bank reserves to boot, doesn’t really need anyone else’s gold (though certain Russian interests might), but that doesn’t mean Russia wasn’t using Sudan’s, with the collusion of certain Sudanese actors. In that respect, the story posed a valid question. Where the piece unraveled was its treatment of actual Sudanese production statistics.
At one point in the video version of the story, reporter Nima Elbagir comments on a spreadsheet provided by a source in the Central Bank of Sudan, indicating that, of 2021’s official total production of 49.7 tonnes, 32.7 tonnes were unaccounted for. It’s certainly possible that over 32 tonnes of gold were smuggled out of Sudan that year – much of it probably to Dubai, directly or indirectly. But that spreadsheet’s accuracy isn’t clear, especially given that 49.7 tonnes is a little over half what was generally accepted by industry bodies such as the World Gold Council as Sudan’s production in 2021. Furthermore, it’s not clear how much of the gold that exited Sudan was mined in Sudan as opposed to trafficked through the country from its gold-producing neighbors. Still, despite those questions, the estimate of missing gold is reasonable.
What isn’t reasonable is what follows: Elbagir asserts that, according to multiple “whistleblower” sources, the actual figure for gold smuggled out of Sudan in 2021 may be over 220 tonnes – 223 tonnes, worth around USD 13.4 billion, if the report’s visual is to be believed. That number would vault Sudan to being not only the biggest gold producer by far in Africa but the fourth-largest on the planet, behind only China, Russia, and Australia – all countries with extensive mining infrastructure and massive large-scale mining operations. And such a scale of production would be impossible to conceal; it would involve industrial operations consuming a great deal of energy, lots of heavy machinery, fleets of trucks, and enormous amounts of waste material. A rough calculation based on reserves at the Hassai mine, Sudan’s most prominent, with ore at a grade of 2.5 g/t, would mean excavating, moving, and processing 89.2 million tonnes of “run of mine” material.
It didn’t take long for Sudan’s Ministry of Minerals to make precisely those points in an open letter rebutting the CNN report. Using a slightly higher average grade of 3 g/t, the Ministry arrived at 92 million tonnes of material to be extracted, moved, and processed – and then detailed how other infrastructure and logistical considerations also revealed that the estimate of 223 tonnes in one year was untenable. We have seen no indication that CNN responded to that letter.
It’s possible that the figure of 223 tonnes in the CNN report represented losses over several years, not just 2021, and somehow the error went uncaught as the report went through production. But the problem here was more fundamental than one errant data point. To all appearances, the story’s reporters and producers didn’t know enough about the subject matter of gold flows to question whether the numbers they were citing even made sense.
It matters little that the written version of the story qualifies its use of those figures, noting that CNN could not “independently verify” them. The error in the video was allowed to continue circulating without correction or clarification. As CNN subsequently reported, the story led to protests on the streets of Khartoum. People risked their lives. Some may have lost their lives. And how many of them took that risk because they were outraged by a statistic that was an unforced error? An error that could have been avoided had CNN’s reporters done their homework and realized that their numbers didn’t match up to how gold supply chains actually work? And how much was the cause of democracy in Sudan set back when CNN’s otherwise strong reporting was so easily discredited because of that error?
Facts without understanding, data without subject matter expertise, can lead to dangerous real-world consequences. This is not to say that analysis can’t go beyond hard facts. It should. A degree of informed, expert speculation is part of the process. But those words “informed” and “expert” are crucial. You have to do your homework, and consult with experts even when you’ve gained expertise of your own. You have to continually revise your map of the terrain. That process never ends, because mastery is a moving target: the more you learn, the more complexities you can see and appreciate and factor into your assessments.
Data must be paired with foundational knowledge – not of data, but of what the data represent. When analysts in any field fail to gain, or at least to access, subject matter expertise, they will make heedless mistakes. And if those errors happen at the wrong time, in the wrong place, people might die in the streets because someone didn’t know what they were talking about.