Causal effect of video game play on mental well-being: a quasi-experimental study among Japanese population

Abstract

The number of people who play video games has become close to three billion during the COVID-19 pandemic. Yet, video games' causal effect on well-being has been understudied, and most studies have relied on correlation. We select k6 and SWLS for measuring well-being and apply causal inference methods. We take advantage of a natural experiment where lotteries for purchasing video game consoles---the Nintendo Switch and the PlayStation 5---created a random variation in access to video games. Using online surveys with 100,000 responses from Japanese aged 10-69 during the COVID-19 pandemic, we find that video game play impacts users' well-being positively. Winning a game console lottery positively affects well-being by 0.2 SD for the Nintendo Switch and 0.1-0.2 SD for the PlayStation 5. The instrumental variable approach indicates that owning a new game console has a positive impact on well-being by 0.6 SD for the Nintendo Switch and 0.1-0.2 SD for the PlayStation 5. In addition, a machine learning method reveals a striking difference between the heterogeneous effects of the two video game consoles. While the PlayStation 5 effect is smaller for teenagers, married, non-gamers, and females, such heterogeneity is not found for the Nintendo Switch effect.