Who shapes our daily choices? Society, friends, family, colleagues, or neighbours? A team of economists examined how social interactions influence our decisions and what this means for public policy.
The acclaimed series The Wire, particularly its first season, portrays the clash between law enforcement and gangs in Baltimore. Beyond its realistic and nuanced storytelling, the series stands out for its complex characters, such as the determined policeman McNulty and the criminal with a passion for economics, Stringer Bell. Despite being on opposite sides of the law, both characters share sharp intelligence and adherence to their own codes of honour.
But how can two such similar individuals end up on such divergent paths? Their stories illustrate the profound impact of environment and peers on the choices we make—a reality explored by economists Yves Zenou, Vincent Boucher, Michelle Rendall, and Philip Ushchev.
Invisible, but omnipresent
The phrase "man is a social animal", often attributed to Aristotle, originate from the Greek term ζῷον πολιτικόν which is more accurately translated as "political animal." This idea reflects the belief that humanity’s natural state is to live in community, governed by a set of rules—some more explicit than others. Our behaviours and attitudes are shaped by a common set of norms that organise and structure life in society. These norms vary across different territories. For instance, in Switzerland, reporting a neighbour who breaks the law is a socially accepted and even encouraged act, whereas this is not the case in neighbouring countries.
Some of these norms are codified into laws that punish actions deemed harmful to society, such as criminal acts. Over the past thirty years, while crime rates in the United States have generally been declining, the population behind bars has continued to rise, making it the country with the largest prison population in the world. This phenomenon is closely tied to the "Zero tolerance" policy implemented by Rudy Giuliani during his tenure as mayor of New York City from 1993 to 20011. His approach to crime was straightforward: increased firmness, reduced time between arrest and trial, more frequent checks, and a larger police presence. The heightened risk of incarceration, he argued, would serve as a deterrent to crime.
- 1
The effectiveness of R. Giuliani’s policy is contested by economists Steven Levitt and Stephen Dubner, who attribute the decline in crime to the legalisation of abortion in 1973. According to them, this measure primarily reduced the number of unwanted children—those most likely to become involved in criminal activity—leading, two decades later, to a decrease in the number of delinquents and criminals.
Picture by Photo Spirit on Adobe Stock
For Yves Zenou and his collaborators, this type of repressive policy is not ideal and may even have the opposite effect. A high incarceration rate can inadvertently foster crime by destabilising family structures. The rise in incarceration rates contributes to an increase in single-parent households, with women as the primary caregivers. These women, often facing economic hardship, struggle to provide a stable environment for their children, which in turn increases the likelihood of those children engaging in delinquent behaviour2. According to the researchers, the focus of crime prevention should shift to targeting the most influential individuals in a community, such as gang leaders. This idea is supported by an innovative model that explores the diffusion of behaviours within a population, particularly those behaviours originating from individuals with the "most extreme traits."
- 2
Bezin, E., Verdier, T. and Y. Zenou (2022), “Crime, broken families and punishment,” American Economic Journal: Microeconomics 14(4), 723-760.
Simulating Human Behaviour
In science, a model is a simplified representation of a phenomenon, designed to explain its past, present, or future behaviour. Models can be used to predict future events, such as forecasting the weather. But when it comes to the social sciences, which study individuals, groups, and their environments, how can we predict something as unpredictable as human behaviour?
To date, most empirical models used to study behavior rely on the assumption that individuals are influenced by the average behavior of their peers—a framework known as the linear-in-means model. For instance, a person is more likely to engage in criminal activity if the average criminal behavior is prevalent within their social circle. Similarly, in an academic setting, a student's motivation to study is often shaped by the average performance of their classmates. However, this approach assumes that all peer interactions exert an equal influence, which can be unrealistic in large groups. The authors extend this model by allowing an individual's actions—such as the number of crimes committed—to be shaped not only by the average behavior of their peers but also by the specific actions of any individual within their network, including those exhibiting the highest, lowest, or any other level of behavior.
The Power of Friendship
To test this more realistic model, the researchers analysed the social interactions of participants in the U.S. AddHealth program. In 1994-1995, individuals aged 12 to 18 were regularly asked to report their friends. The analysis revealed that each person is influenced by both the "best" and "worst" behaviours of their peers, particularly the "most extreme" behaviours within their social circle. For example, being close to a high-achieving student encourages you to emulate their behaviour and fosters greater engagement in studies. Conversely, having a friend who engages in risky behaviours, such as extreme sports, can lead to distancing yourself from them.
Targeting nodes of influence
The series The Wire introduces characters like Wallace and Bodie, young individuals who become entrenched in the cycle of crime. They aspire to imitate those they perceive as "successful." By highlighting the influence and spread of behaviours from the most "extreme" individuals, the work of Yves Zenou and his colleagues is paving the way for new public policies aimed at key players.
Their research demonstrates how the actions of the most extreme individuals—often leaders—can propagate throughout a community. This insight offers a fresh perspective on how public policies can be designed to target influential figures, such as gang leaders, in order to more effectively combat crime.
A common approach in anti-crime policies is to increase police presence and surveillance in areas known for criminal activity. However, based on their research, Yves Zenou and his colleagues suggest a more targeted strategy: focusing on the specific locations where criminals gather and plan their actions. This approach would not only reduce crime but also optimise resource allocation. According to the researchers, this strategy could save a city like London approximately 300 million euros per year3.
Much like in The Wire, where police officers focus on the criminal network's "nodes of influence"—strategic locations where criminals meet to organise their activities or hold secret meetings—the idea is to concentrate efforts on the areas where influence is most concentrated, rather than spreading resources too thin across broader territories.
- 3
Giulietti, C., McConnell, B. and Y. Zenou (2015), “Hot Knots vs Hot Spots: Targeting Key Players in Crime,” Unpublished manuscript, Monash University.
