Have we vastly underestimated the total number of people on Earth?

Our estimates of rural populations have systematically underestimated the actual number of people living in these regions by at least half, researchers have claimed – with potentially huge impacts on global population levels and planning for public services. However, the findings are disputed by demographers, who say any such underestimates are unlikely to alter national or global head counts.

Josias Láng-Ritter and his colleagues at Aalto University, Finland, were working to understand the extent to which dam construction projects caused people to be resettled, but while estimating populations, they kept getting vastly different numbers to official statistics.

To investigate, they used data on 307 dam projects in 35 countries, including China, Brazil, Australia and Poland, all completed between 1980 and 2010, taking the number of people reported as resettled in each case as the population in that area prior to displacement. They then cross-checked these numbers against five major population datasets that break down areas into a grid of squares and estimate the number of people living in each square to arrive at totals.

Láng-Ritter and his colleagues found what they say are clear discrepancies. According to their analysis, the most accurate estimates undercounted the real number of people by 53 per cent on average, while the worst was 84 per cent out. “We were very surprised to see how large this underrepresentation is,” he says.

While the official UN estimate for the global population is around 8.2 billion, Láng-Ritter says their analysis shows it is probably much higher, though declined to give a specific figure. “We can say that nowadays, population estimates are likely conservative accounting, and we have reason to believe there are significantly more than these 8 billion people,” he says.

The team suggests these counting errors occur because census data in rural areas is often incomplete or unreliable and population estimation methods have historically been designed for best accuracy in urban areas. Correcting these systematic biases is important to ensure rural communities avoid inequalities, the researchers suggest. This could be done by improving censuses in such areas and recalibrating population models.

If rural population estimates are way out, that could have massive ramifications for the delivery of government services and planning, says Láng-Ritter. “The impacts may be quite huge, because these datasets are used for very many different kinds of actions,” he explains. This includes planning transport infrastructure, building healthcare facilities and risk reduction efforts in natural disasters and epidemics.

But not everyone is convinced by the new estimates. “The study suggests that regional population counts of where people are living within countries have been estimated incorrectly, though it is less clear that this would necessarily imply that national estimates of the country are wrong,” says Martin Kolk at Stockholm University, Sweden.

Andrew Tatem at the University of Southampton, UK, oversees WorldPop, one of the datasets that the study suggests was undercounting populations by 53 per cent. He says that grid-level population estimates are based on combining higher-level census estimates with satellite data and modelling, and that the quality of satellite imagery before 2010 is known to make such estimates inaccurate. “The further you go back in time, the more those problems come about,” he says. “I think that’s something that’s well understood.”

Láng-Ritter thinks that data quality is still an issue, hence the need for new methods. “It is very unlikely that the data has improved so dramatically within 2010-2020 that the issues we identified are fully solved,” he says.

Stuart Gietel-Basten at the Hong Kong University of Science and Technology points out that the majority of the team’s data comes from China and other parts of Asia, and may not be globally applicable. “I think it’s a very big jump to state that there is a great undercount in places like Finland, Australia, Sweden etc, and other places with very sophisticated registration systems, based on one or two data points.”

Láng-Ritter acknowledges this limitation, but stands by the work. “Since the countries that we looked at are so different, and also the rural areas that we investigated have very different properties, we’re quite confident that it gives a representative sample for the whole globe.”

Despite some reservations, Gietel-Basten agrees with Láng-Ritter on one point. “I certainly agree with the conclusions that we should both invest more in data collection in rural areas as well as coming up with more innovative ways of counting people,” he says.

But that idea that the official world population should swell by a few billion “is not realistic”, says Gietel-Basten. Tatem also requires much more convincing. “If we really are undercounting by that massive amount, it’s a massive news story and goes against all the years of thousands of other datasets” he says.