Headlines grossly misrepresented a groundbreaking paper through misinterpretation, miscalculation, reproduction and automation. The original researcher explains why this is so damaging.
In the grand scheme of environmental fears, few create more of an existential sense of sealed fate than talk of the planet running dry. Oxygen may be fundamental to the survival of many species, but no known terrestrial form of life can exist without a source of water.
So, in March, when the Northern Hemisphere was staring down the calendar at another dry summer of heatwaves, wildfires and associated deaths, news that there could be up to 2 billion more humans in need of dwindling supplies hit hard. Very hard.
‘There could be billions more people on Earth’ warned The Independent. ‘Have we vastly underestimated the total number of people?’ asked New Scientist. While The Guardian led with: ‘the world’s population stands at just over 8.2 billion. However, a recent study suggests the figure could be hundreds of millions or even billions higher.’ Environment Journal followed suit, albeit with slightly less drama: ‘scientists may have miscalculated how many people are on Earth.’
Nevertheless, the story across hundreds of publications was on a similar tip: the planet’s human population could be one-third higher than previous estimates. But this isn’t actually the case. Josias Láng-Ritter, a PhD researcher specialising in ‘the impacts of society on water, and vice versa’ at Aalto University, Finland, led the study in question, which did show that rural population size is up to 50% lower than the reality. Crucially, though, trying to extrapolate this to the total global population is a huge mistake. As a number of news outlets made, catalysing even more re-written stories. Thirsty AI models then slurped this up, before delivering results to users based on accidental misinformation.
‘The study assessed the accuracy of global population maps in rural areas specifically. These maps have been widely used in policy making, for instance in the allocation of resources around the world, but their accuracy in rural areas was largely unknown,’ explains Láng-Ritter. ‘We used as a reference ground-based counts of populations in 35 countries that were displaced due to the construction of dams.
‘When comparing these ground-based population counts to the numbers of people the global population maps showed in these areas, we noticed that the maps systematically underestimated the ground-based counts… [meaning] people living in rural areas,’ he continues. ‘While our results provide a first indication of systematic underrepresentation of rural communities in global population maps, they also contain large uncertainties.’
If one thing is certain, systemic underrepresentation of this kind can lead to population invisibility at policymaking level. So swathes of people are in danger of being left out when it comes to analysis of available and remaining resources. Or how much infrastructure is needed in a specific area. What the research definitely doesn’t say, though, is that there could be up to 2 billion more people on the planet than we thought in January.
‘Some of the big headlines simplified our findings to one main message: There are many more people living on planet Earth than the assumed 8.2 billion,’ says Láng-Ritter, before explaining how the seeds of the misinformation deluge began – rushed maths conducted by people who misunderstood how to crunch these kind of numbers. ‘Some media even made their own flawed calculations to quantify the missing number of people to 2-3 billion.
‘Such a large number of missing people is absolutely out of the question,’ he continues. ‘For example, one possible reason for the underestimation of rural population in global maps could be that they are counted as belonging to different locations, e.g. in urban areas,’ he continues. So people might have been counted, but the maps show them in the wrong place, so to speak.’
This is one story, and the fallout was significant. Our heart goes out to the university’s comms department, and the understandably frustrated International Editor who had to contact countless publications – including ourselves – to inform them of inaccuracies. But at a moment in time when the biggest risk to publications that make mistakes is no longer just declining trust, context is more troubling than the example. A headline with weight can rapidly produce more stories, written by humans and machines, automatically spawning many more. Inaccuracies quickly become more visible than the truth.
‘[This] is very damaging. I think media misrepresentations can compromise the trust that society puts into research, which in the long-term can lead to decision making that is less based on evidence and more on emotions. It can be very damaging to the credibility of research and researchers if science news is simplified into catchy headlines and the complexities and limitations not given a mention,’ warns Láng-Ritter.
‘I don’t think I would be more cautious about publishing potentially sensitive research as long as it raises important questions for our society,’ he continues, before moving t what he would do differently next time, and why change could be an issue in itself. ‘I would try to point out more clearly and early on the limitations of the research, to further reduce the risk of the media jumping to their own conclusions without contacting us for comments. But this strategy may bring the side effect that important research does not end up reaching the audience it deserves. It’s a fine line.’
More on Misinformation:
How to spot climate misinformation and disinformation, and counter both
Stricter rules needed to stop MPs repeating conspiracy theories