• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

Everybody Ran and I Ran, Too

Why do we tend to violate the lockdown rules? The answer lies in the game theory.

© gorodfinansov.ru

In lockdowns, why do some people stay home, while others violate the quarantine rules and go out for picnics in the park? Behavioural economics may provide the answer to this question. Oksana Zinchenko, a Research Fellow of the Institute of Cognitive Neuroscience, explains how we can predict people’s behaviour with game theory.

Oksana Zinchenko
Research Fellow Institute of Cognitive Neuroscience, HSE University

When we started hearing recommendations to self-isolate, before the administrative sanctions and penalties were imposed, we came across a new social norm—many people (with their employer’s approval) started working from home voluntarily. They began to leave home as little as possible, ordering things online instead of going shopping. However, along with this socially responsible behaviour and care for the common good, we also observed a multitude of counter examples of people going out on picnics to parks on warm days at the beginning of April, despite the climbing numbers of confirmed COVID-19 cases and calls for people to stay home. Why do some people follow new social norms without the threat of penalties, while others regularly break them?

In behavioural economics, group behaviour is studied by using game theory models, such as public goods games. In a standard public goods game, each subject, or individual player, has his or her own amount of tokens. The player secretly chooses how many of them to put into the public pot (pool) in order to set up a shared budget.

After each round, the tokens in the pot are multiplied by a factor (greater than one and less than the number of players) and the payoff is divided evenly among the players. The group's total payoff is maximized if everyone cooperates; if not, the most rational behaviour would be to contribute zero and expect to benefit from other players’ contributions.

A large volume of data has been accumulated to date on the dynamics of people’s group behaviour by using this model in situational tests. What can we infer about the current situation based on this data?

The level of contribution to the common good is known to be rather high at the beginning, but it tends to get lower after some time. Experiments show that about 60–80% of the group members reduce the amount of their contribution to the pot — thus they stop supporting the social norm. The others contribute much smaller amounts in later rounds compared to their contributions at the beginning of the game. Looking at the self-isolation patterns in Russia generated by Yandex, we can see a similar downturn trend. Data for the city of Moscow suggests that this trend emerged later.

Why is this the case? Behavioural economists Simon Gächter, Urs Fischbacher, and Ernst Fehr have proved that, on average, about half of players are ‘conditionally cooperative’ — they watch the other group member’s behaviour and stick to the norm only if they see that the other beneficiaries are willing to contribute. 

About another third of players are highly motivated by private benefits, ignoring the norm (not contributing to the public pool). Therefore, the conditional cooperators may observe a gradual increase in instances of others not sticking to the norm and tend to act accordingly — they stop adhering to the norm themselves, contributing less and less to the public pool.

Public goods games are widely used to model scenarios of environmental pollution and using non-renewable natural resources. The behavioural financier Colin Camerer says that players who are focused primarily on their own benefits avoid contributing to the public pool (refusing to stick to the norm) regardless of what other players do. 

However, this behaviour can be effectively changed. One of the ways to do so is for the government to impose administrative measures. This could be in the form of fines and penalties or special taxes, for instance, to prevent environmental pollution. The other way appears to be unofficial enforcement mechanisms, which would require actions from other members of the social group or society at large. Such actions include social condemnation and ostracism.

An experiment conducted by Ernst Fehr and Simon Gächter in 2000 shows that the opportunity for other members of a social group to punish those who deviate from the group standard (or those unwilling to contribute to the public pool) causes a large increase in the cooperation levels.

Note: The above pattern shows the results of the experiment, during which the group members interacted with the same players inside their group at each game round. However, the group in which members changed before the beginning of every round demonstrated a similar dynamic.

IQ

Author: Oksana Zinchenko, May 31