Mobile payments and crime: evidence from China

ABSTRACT

Using a policy change in 2016 as a natural experiment, we employ a Difference-in-Difference (DID) model to analyse the effects of mobile payments development on crime rates in Chinese prefectures from 2015 to 2019. Our findings indicate that mobile payments expansion has a significant negative effect on theft, with the reduction of residents’ cash holdings serving as a potential mechanism. However, we find no significant impact on non-economic crimes like sexual assault and murder. This study provides evidence supporting the role of policies and technologies enabling mobile payments development in deterring crime and enhancing social security.

KEYWORDS:

• Mobile payments
• crime
• cash
• DID method
• DDD method

JEL CLASSIFICATION:

• E42
• G23
• K24

Disclosure statement

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

Notes

¹ China Judgments Online was established by the Supreme People’s Court of China in 2016, which is the official online platform that publishes all legal judgements.

² China Judgements Online is accessible at https://wenshu.court.gov.cn/. The retrieval process of the theft case data in this paper can be found in the Table B1 of Appendix.

³ The criteria for prosecuting theft crimes underwent four revisions, occurring in 1984, 1992, 1998, and 2013, but these adjustments do not affect our sample data. While we acknowledge the existence of unreported criminal cases, often referred to as ‘black numbers’, it’s important to note that these cases are typically relatively minor (Skogan 1977). Therefore, using official records remains a reliable method for measuring the incidence rate (Levitt 1998).

⁴ According to Liang and Jiang (2020), the location of the trial court can reflect the geographical distribution of theft cases.

⁵ In the robustness test of the empirical results, we discuss the plausibility of lagging our data by one period.

⁶ The complete list of cities in our sample can be found in the Table A1 of Appendix.

⁷ The policy shock in 2016 is aligned with crime data from 2017; hence, the post indicator corresponds to 2018 and 2019.

⁸ The policy shock in 2016 is aligned with crime data from 2017.

⁹ The retrieval process of the sexual assault case and murder case data in this paper can be found in the Table B1 of Appendix.

¹⁰ The data obtained from the question ‘How much cash does your household currently hold’ in the CHFS 2015 and 2017 questionnaires.

¹¹ The question regarding risk attitude in the survey: If you have a choice, what kind of investment will you make? 1. High risk and high return projects; 2. Secondary high-risk and secondary high-return projects; 3. Average risk and average return projects; 4. Secondary low-risk and secondary low-return projects; 5. Unwilling to take any risks. This paper defines option 3 as risk neutral, while option 1 and option 2 are defined as risk preference, and option 4 and option 5 are defined as risk aversion.

¹² The standard deviation of payment usage index is 48.81. Because the unit of household cash holdings per capita is 10,000 yuan, we calculate this value by 48.81 × 0.0057 × 10000 = 2782.17.

Additional information

Funding

Jingjing Zhao is grateful to the financial support of State Scholarship Fund of China Scholarship Council number 202208110129. Zongye Huang is grateful to the financial support of Key social science project of Beijing Municipal Education Commission grant number SZ202110038017.