Loot box gambling and economic preferences: a survey analysis of Japanese adolescents and young adults

ABSTRACT

With the increasing use of social-network games, game addiction has become a serious challenge to the world. This study investigates the effects of risk preference, loss aversion, and time preference on the behaviour of Japanese adolescents and young adults purchasing ‘Gacha’, or loot box gambling, in social-network games. We surveyed 1,210 respondents, aged 12 to 23 years; approximately 34% of the respondents had purchased ‘Gacha’. We found that loss-averse and risk-averse female respondents had less experience paying for ‘Gacha’, with their highest billing amount charged per month being lower than that of other female respondents. Furthermore, future-oriented female respondents had less payment experience than present-oriented ones, with the highest billing amount charged per month for the former respondents being also lower. The highest billing amount charged per month of loss-averse male respondents was significantly lower than that of other male respondents.

KEYWORDS:

• Gacha(loot box gambling)
• social-network games
• adolescents and young adults
• risk preference
• time preference

JEL CLASSIFICATION:

• D12
• D91

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ According to Koyama (2016, 324), loot box gambling refers to a system of purchasing characters and virtual items by drawing lots in social-network games on smartphones and mobile phones. The type of character or virtual item that the user receives depends on probability. In Japan, loot box gambling is called “Gacha”. This name comes from the term “Gacha” or “Gacha-Gacha”, which refers to “Gashapon” coin-operated toy vending machines that release a toy in a plastic capsule. Since social network games often have effects that imitate the rotation of the “Gacha-Gacha”, players call the loot box “Gacha-Gacha”. In Japan, playing a single “Gacha” costs from JPY 100 to 500, with the cost of playing “Gacha” 10 times usually being sufficient to draw 11 raffles.

² In the Japanese market, only Pokémon Go does not have a “Gacha” for obtaining a character.

³ JOGA’s guidelines are shown at https://japanonlinegame.org/joga_guideline/joga-guideline/, CESA’s guidelines ares shown at https://www.cesa.or.jp/guideline/social.htm..

⁴ See Kahneman (2011)..

⁵ In most cases, charging social games is equal to purchasing “Gacha”.

⁶ Sutter, Zoller, and Glätzle-Rützler (2019) provided an excellent survey of experimental studies on youth.

⁷ Castillo et al. (2011) estimated the time preference of students aged 13 and 14 years in Georgia and found that the number of discipline referrals was larger for those with a high time discount rate compared to other students. Castillo, Jordan, and Petrie (forthcoming, a follow-up study of Castillo et al. (2011), showed that adolescents with higher time discount rates had lower high school graduation rates. Castillo, Jordan, and Petrie (2018) showed that students with lower risk aversion at the ages of 13–14 years had more disciplinary referrals and lower high school graduation rates. Golsteyn, Grönqvist, and Lindahl (2014) pointed out that time preference at the age of 13 years predicts long-term outcomes such as school performance, lifetime earnings, and unemployment. Sutter et al. (2013) measured time, risk, and ambiguity preferences for 10–18-year-olds in Austria and examined their correlation with various behaviours. Their results showed that adolescents with higher time discount rates were more likely to smoke and drink, have a higher body mass index (BMI), save less, and behave poorly in school.

⁸ The brief results for Schools A, B, and C are reported in Ogawa et al. (2019).

⁹ Some users may think of the loss as a situation wherein he or she cannot complete a series of items or characters and considers that there is a further opportunity for offsetting this loss by obtaining more “Gacha”. Those who think in this way are likely to be immersed in social network games and consequently are charged more. As the number of respondents with high billing amounts was small in our survey, we did not consider this possibility.

¹⁰ In Booth and Nolen (2012), the risk preference question was the “Fiver Lottery”. In this lottery, “each student chooses Option 1 or Option 2. Option 1 is to obtain $=$ for certain. Option 2 is to flip a coin and obtain undefined if the coin comes up heads or ${ }^{\ast }$ if the coin comes up tails”.

¹¹ Camerer and Hogarth (1999) pointed out that whether a participant is given monetary incentive does not affect the mean performance in risky choices. Further, Holt and Laury (2002) found no significant difference between hypothetical and real choices.

¹² In Moreira, Matsushita, and Da Silva (2010), participants aged four to six years were offered the choice between a visible quantity of 150 ml of strawberry juice and random receipt of 0 or 300 ml of juice. Depending on the outcome of their choice, they could obtain juice(0, 150, or 300 ml). The results showed that the children were risk-seeking.

¹³ If the number of naive respondents was large, they might have more experience of purchasing “Gacha”, with large highest amounts of money spent purchasing it in a month. Meanwhile, we assumed that the number of sophisticated respondents is high. One type of sophisticated respondents decides to purchase “Gacha” and actually purchases it, whereas the other type decides not to purchase it and accordingly does not do so.

Additional information

Funding

This study was supported by the Japan Society for the Promotion of Science KAKENHI (grant number 19K01769), the Regional Collaborative Projects Fund at Kansai University, and the Publicly Offered Research Grant of Research Institute for Socio-network Strategies at Kansai University. We sincerely thank all the elementary, middle, high school, and university students who participated in the survey. We also thank students from the Ogawa and Motonishi seminars at Kansai University for their cooperation in conducting the surveys.