Predicting the payment preference for CBDC: a discrete choice experiment
Summary
Focus
We examine how people in Korea may respond to the introduction of a retail central bank digital currency as a payment method. Since no real usage data exist, we conduct a discrete choice experiment with more than 3,500 individuals, where participants evaluate pairs of hypothetical payment methods defined by attributes such as issuer, form of issuance, fees, timing of settlement and discounts.
Contribution
The experiment design allows us to identify the role of each attribute in shaping payment choices and to predict demand for a retail central bank digital currency under different design scenarios. By validating the predictions against actual usage of existing payment methods in Korea, we show that this approach provides credible and reliable evidence for policy discussion.
Findings
We find that financial incentives, such as fees and discounts, have the strongest influence on payment choices, and issuer and form of issuance also matter significantly. Respondents show a substantially higher willingness to adopt payment methods when they are provided in the form of cards or smartphone apps, rather than banknotes. Also, older respondents in particular show greater trust in central bank-issued instruments, while younger respondents express stronger preferences for big tech and app-based options. In simulations, retail central bank digital currency with a benchmark design is chosen as the most preferred payment method by 19% of respondents, ranking second to cards but ahead of cash and mobile payments. Predicted use increases further when it provides stronger financial incentives, highlighting that design features – including incentives, issuer credibility and issuance form – will critically shape its acceptance.
Abstract
To overcome the lack of data in predicting the payment preference for central bank digital currency (CBDC), we conducted a discrete choice experiment that varied the attributes of payment methods among over 3,500 participants in Korea. We identified key attributes, such as the discount rate and the issuance form, that shape the demand for payment methods. The predicted usage shares of existing payment methods closely align with their actual usage patterns in Korea, which lends credible support for the external validity of our experimental design. Building on this validation, we further predict that CBDC, when introduced, will be preferred over cash and mobile fast payment but less preferred than credit and debit cards, with its adoption rate as the most preferred payment method ranging 19−27% of respondents.
JEL classification: E40, E50, C90
Keywords: payment preference, retail CBDC, discrete choice experiment