Pervasive Mobile Payment Adoption and the Decline of Cashiers’ Mathematical Skills
Description
This data set helps to investigate how noncash payment affect the rounding up probability of the next cash transaction, a proxy for cashier's mathematical skills, as well as its impact on transaction time of rounded up transactions that require calculations. The files "main.dta" and "main.do" reveal the main finding. The greater the number of consecutive noncash transactions, the lower the probability of rounding up for the next cash transaction. The files "round up possibility.dta" and "round up possibility.do" apply another approach to support this argument. I construct pairs of transactions that have N (N=2, 3, ..., 10) consecutive cash transactions (noncash=0) or noncash transactions (noncash=1) before. For example, N3=1 means that for this transaction of interest, there are three noncash transactions before it, vice versa, N3=0 means that there are three cash transactions before it. Comparing in pairs with consecutive cash transactions, noncash payment series lowers the probability of rounding up of the next cash transaction. The files "transaction time.dta" and "transaction time.do" show the time-increasing effect of noncash transactions on the following rounded up transaction by comparing the unit transaction time in pairs. roundup is a binary variable. It equals to 1 if the transaction is rounded up. noncashseries is the number of consecutive noncash payments. service is the number of consecutive services of a cashier in his or her shift. item is the number of products purchased. One may argue that the number of cash or noncash series is endogenous as cashiers may intentionally imply customers to pay in cash or noncash. Therefore, we use a Bartik-type IV, iv, defined as the cash proportion of the whole store before a transaction. Individual cashier has no power to manipulate the population cash-using proportion. The cash proportion affects the noncashseries but has no effect on round up probability. peakhours, weekends and holidays are subsamples to capture a relatively busy working conditions.