New study reveals that city isolation, not size, may drive political violence in Africa.
Cities are often seen as hotspots of violence, with the assumption that larger cities are inherently more violent than smaller ones. This “universal law” of urban scaling has long shaped scientific thinking. But new research led by Complexity Science Hub (CSH) researcher Rafael Prieto-Curiel challenges this assumption. Published in Nature Communications, the study shows that it is not simply city size, but a city’s level of isolation, that plays a crucial role in determining violence.
“Our analysis shows that the 10% most populous cities in Africa (216 cities in total) contain 66% of the urban population but only 33% of the fatalities linked to politically-motivated violence over the past 22 years,” explains Prieto-Curiel. “This clearly indicates that size alone cannot explain levels of urban violence.”
WHEN ISOLATION FUELS VIOLENCE
Coming from Mexico, Prieto-Curiel recalled Aguililla, a cartel stronghold connected to the rest of the country by only a single highway. “It’s relatively easy for cartels to know when someone – say, from the police – is approaching,” he explains. This led him to investigate whether a city’s “level of isolation,” defined as the number of highways connecting it to other cities, influences the extent of political violence in general.
Using OpenStreetMap data, Prieto-Curiel constructed a network of all highways in Africa. The results show: cities with only one or two highway connections experienced nearly seven times more violence against civilians per 100,000 residents than cities with seven or more highways. Comparing the 10% most isolated cities with the 10% least isolated, the researchers found that the former faced more than seven times the level of violence against civilians. Moreover, looking at the top 25% most and least isolated cities from 2000 to 2020 revealed that the gap has widened over time. Isolated cities are becoming more violent in recent years.
ISOLATION WITHIN ISOLATION
To refine their analysis, Prieto-Curiel, together with his colleague Ronaldo Menezes from the University of Exeter, introduced a second measure of isolation: city centrality.
This captures not only how many highways lead to a city, but also how connected it is within the broader network of cities. “Imagine a smaller city next to a large, well-connected city – like London, Cairo or Jakarta, for example.
Even if the smaller city only has a single highway, people can still reach all of the bigger city’s highways and facilities, such as hospitals, quickly and easily.” But if an “isolated” city is surrounded by other equally isolated cities with only one or two highways, the situation is very different.
To capture this, the researchers built a model that detected the shortest route for residents of every city in Africa to every other city on the continent and simulated millions of those journeys, then measured how many of them pass through a given city. “If hardly anyone passes through a city, then it is very isolated,” explains Prieto-Curiel.
Again, the findings showed a similar picture – for example, in cities ranking in the lowest 25% for city centrality, lethality rates were 15 times higher than in the 25% with highest city centrality.
“We needed a new way of understanding why smaller, more remote African cities experience disproportionately higher levels of political violence. My background in network science and crime analysis, along with my experience working with police forces in Brazil and the UK, helped map the relationship between urban isolation and violence intensity in African cities,” Menezes says. “We developed a methodology that measures city isolation through highway connections and travel patterns, which revealed isolated cities face four times higher casualty rates than well-connected urban centres. Our research has given vital insights into potential interventions such as targeted security and infrastructure planning.”
DATA ON POLITICAL VIOLENCE
To study violence, the researchers relied on data from ACLED (Armed Conflict Location & Event Data Project), which compiles media reports of armed conflicts worldwide.
For this study, they analyzed nearly 300,000 politically-motivated events in Africa between 2000 and 2023, linked to almost 600,000 casualties. These were categorized as battles (25%), violence against civilians (25%; where an organized armed group deliberately inflicts violence upon unarmed non-combatants), protests (24%), riots (11%), explosions (8%), and strategic developments (7%).
While not perfect, Prieto-Curiel stresses, “it is as close as you can get” to a comprehensive geography of conflict in Africa. He also notes that violence in small, remote cities is likely underreported, meaning the disparities may be even greater than observed.
URBAN HETEROGENEITY
The research highlights the importance of moving beyond universal laws when studying cities. “We need to be very conscious of how heterogeneous the world is. We cannot observe cities as a universal phenomenon that can be explained with fixed physical laws,” says Prieto-Curiel. “Big cities are not inherently violent.”
About the study
The study “Violence, City Size and Geographical Isolation in African Cities” by Rafael Prieto-Curiel and Ronaldo Menezes was published in Nature Communications (doi: 10.1038/s41467-025-65728-6).
