The Rules of Contagion

Adam Kucharski argues that outbreaks of all kinds, from infectious diseases and financial crises to online misinformation and gun violence, follow identifiable rules of contagion. By uncovering these rules, he contends, we can better predict, measure, and control what spreads.

Kucharski opens with a personal account of his research group's real-time analysis of the early COVID-19 pandemic in February 2020, illustrating the difficulty of interpreting contradictory data during a fast-moving outbreak. He describes key early findings, including evidence that infected people could spread the virus before showing symptoms and an estimated infection fatality rate (the proportion of infected people who die) of about 0.5 percent, derived from international data including testing aboard the Diamond Princess cruise ship. As undetected outbreaks surfaced in Iran, Italy, and the UK, governments faced divergent choices about how to respond. The pandemic also unleashed other forms of contagion: misinformation, economic disruption, and cyber-attacks. This frames the book's central argument that outbreaks of many kinds follow the same pattern of spark, growth, peak, and decline, and that principles governing one type can illuminate others.

Kucharski traces the intellectual origins of outbreak science to Ronald Ross, a British army surgeon who discovered in the 1890s that mosquitoes transmit malaria. Ross developed a mathematical model showing that malaria could be controlled without eliminating every mosquito; below a critical density, the disease would fade on its own. This "mechanistic" approach, which starts with assumptions about how transmission works and uses equations to predict outcomes, differs from the "descriptive" method of finding patterns in existing data. Kucharski argues that mechanistic models allow researchers to ask "what if" questions essential for evaluating interventions. He introduces William Kermack, a chemist who lost his sight in a laboratory accident, and Anderson McKendrick, who had worked with Ross. In the 1920s they developed the SIR (Susceptible-Infectious-Recovered) model, which showed that epidemics can end before everyone is infected because the shrinking pool of susceptible people causes recoveries to outpace new infections. This insight underpins the concept of herd immunity: When enough people are immune, the population as a whole blocks transmission. Kucharski applies these models to outbreaks of Zika virus, which had been linked to birth defects, in the Pacific and Caribbean, showing how researchers used them to estimate transmission rates and plan for medical resources.

Broadening Ross's original vision, Kucharski describes Ross's proposed "Theory of Happenings," which covered events beyond disease. Ross distinguished between "independent happenings," where one person's experience does not affect another's, and "dependent happenings," where contagion drives spread. Dependent happenings produce S-shaped growth curves, a pattern later popularized in sociology by Everett Rogers and applied to marketing by researcher Frank Bass.

Kucharski draws parallels between disease epidemics and financial crises. He introduces the reproduction number (R), the average number of new cases generated by a single infected person, and breaks it into four components he calls "DOTS": Duration of infectiousness, Opportunities for transmission, Transmission probability, and Susceptibility. Different control measures target different components: condoms reduce transmission probability for HIV, masks do the same for COVID-19, and vaccination reduces susceptibility. He explains superspreading, where a small fraction of cases drives most transmission, and warns against blaming specific individuals. The false designation of Gaëtan Dugas as HIV "patient zero" in North America, based on a misread letter in a public health investigation, illustrates how blame narratives distort understanding. After the 2008 financial crisis, ecologists and epidemiologists, including Robert May and Bank of England economist Andy Haldane, applied epidemic models to the banking system. The pre-crisis network exhibited features that amplify contagion: a few dominant hubs, hidden loops, and shrinking distances between institutions. Post-crisis reforms included requiring banks to hold more capital based on their network importance and routing derivatives through central hubs.

The book examines how social contacts shape transmission. Kucharski explains how children's high contact rates drive flu pandemics and how the 2009 UK pandemic's two-peak pattern resulted from summer school holidays interrupting transmission. He tackles the difficulty of proving social contagion, distinguishing it from homophily (the tendency to choose friends with similar traits) and shared environmental factors. Physician Nicholas Christakis and social scientist James Fowler's studies using the Framingham Heart Study, a long-running cardiovascular health study, suggested that obesity, smoking, and happiness could spread through social networks, but the work drew criticism for not ruling out alternative explanations. Kucharski introduces "complex contagion," where people need multiple exposures before adopting a behavior, unlike simple disease transmission.

Kucharski argues that violence can be understood as a contagious phenomenon. Gary Slutkin, a physician who spent a decade fighting disease epidemics in Africa, recognized familiar outbreak patterns in Chicago's gun violence and founded Cure Violence, a program that hires "violence interrupters" from affected communities to break chains of retaliation. Sociologist Andrew Papachristos estimated a reproduction number of about 0.63 for gun violence in Chicago, with fewer than 10 percent of shootings leading to 80 percent of follow-up attacks. Kucharski discusses Florence Nightingale, whose pioneering use of statistics and data visualization persuaded authorities to improve hygiene in Crimean War hospitals, as an early example of using data for public health advocacy. He examines the US opioid crisis as a slow-burning epidemic and discusses how predictive policing algorithms can perpetuate bias by reflecting existing patrol patterns rather than actual crime distribution.

Turning to online contagion, Kucharski challenges the assumption that content "goes viral" in the epidemiological sense. He traces Jonah Peretti's experiments with online content, from a 2001 email exchange with Nike to the founding of BuzzFeed, and examines sociologist Duncan Watts's research showing that hidden influencers do not reliably spark massive outbreaks. About 95 percent of Twitter posts are never shared, and most large cascades are driven by a single broadcast event rather than person-to-person transmission. False news spreads further and faster than true news, driven by novelty, and "information laundering," in which fringe groups target journalists and politicians, allows marginal ideas to reach mainstream audiences.

Kucharski explores how malware exploits computer networks, from the first computer virus in 1982 to the 2016 Mirai attack, in which a botnet (a network of infected devices controlled remotely as a group) overwhelmed a key internet infrastructure provider and took down major websites. Mirai originated not from state-sponsored hacking but from a college student's scheme to attack rival Minecraft game servers. The extreme variability in network connectivity allows even weak viruses to persist through superspreading events at highly connected hubs, and malware evolves to evade detection, paralleling influenza's evolutionary arms race with the human immune system.

The book examines how phylogenetic analysis, which traces evolutionary relationships between organisms through their genetic sequences, has become a tool for tracking outbreaks. Researchers used such methods to trace SARS from bats through civets to humans, track COVID-19 in Washington State, and identify sexual transmission of Ebola from a recovered survivor more than a year after infection. Kucharski extends the evolutionary framework to culture, describing how anthropologist Jamie Tehrani used phylogenetic methods to analyze folktales like "Little Red Riding Hood," tracing some stories back over 4,000 years. He also raises privacy concerns arising from the growing availability of personal data, from the identification of individuals in "anonymized" medical records to GPS data brokers selling movement information.

Kucharski concludes by examining the tension between data-driven research and privacy. He contrasts his own citizen science project with the BBC, in which tens of thousands of volunteers knowingly contributed data, with Cambridge Analytica's covert harvesting of Facebook data to profile voters. When COVID-19 reached the UK, the BBC dataset proved useful for analyzing contact tracing strategies. He argues that ideas from infectious disease research now help address problems across many domains, and that understanding contagion requires continuously searching for weak links and missing links in chains of transmission.