In fact, the latter period corresponds to the period of PayPay campaigns, which is indicated in the top right-hand panel as the time-series change in the fraction of individuals under such campaigns. [...] We calculate this by searching the region of residence at the municipality level for each individual at the end of 2020 and taking the value of 1 for each individual and week if the individual lives in a municipality under PayPay campaigns in the week, and 0 otherwise. [...] Moreover, this campaign effect is long-lasting: in June 2022, approximately 18 months after the PayPay campaigns, the difference in the fraction of QR code users between the two regions remains: the fraction of users in the campaign region is 0.108, while that outside the region is 0.099. [...] This table also presents the amount of QR code payments per capita, which is the mean or median of the amount of QR code payments in the four months. [...] A comparison with the positive coefficient for IV suggests that a shift of preference for the methods of transaction from cash to the QR code causes a downward bias in the estimate of γ0 when we use the OLS.
- Pages
- 33
- Published in
- Japan
