IMF Working Papers

For central bank digital currencies (CBDCs) to accomplish their intended objectives, it is necessary for both consumers to use them and for merchants to accept them. This paper develops a dynamic two-sided payments model with both heterogeneous households and merchants/firms to study: (1) The adoption of CBDC by households and firms, and (2) The impact of CBDC issuance on financial inclusion, informality, and disintermediation. Our model shows that there is a feedback loop where more households will adopt CBDC if more firms accept CBDC and vice versa -- incentivizing both households and firms will result in greater levels of take-up. Households are more likely to adopt CBDC if it is low cost, provides an attractive savings vehicle, reduces the cost of remittances, improves the efficiency of government payments, and (if accepted by merchants) offers a valuable means of payment. Firms are more likely to accept CBDC if fees are low, if there are tax exemptions or subsidies for transactions made in CBDC, and if households who prefer to make payments with CBDC make up a large share of revenue. Upon CBDC issuance, an economy can get stuck at a steady state with low CBDC adoption and small welfare gains if the features of CBDC which do not rely on merchant acceptance (remuneration, efficiency of cross border and government payments) are not sufficiently attractive, or if the households benefiting from these features make up a small share of merchant revenue. Temporary subsidies and using CBDC for government payments can spur initial take-up to transition an economy to a welfare improving steady state with high(er) CBDC usage. Greater adoption of CBDC will result in greater financial inclusion and formalization, but potentially the disintermediation of banks and card payments. Thus, there is a trade-off in designing CBDC for greater adoption. However, the gains are more likely to outweigh the risks in lower income economies with larger unbanked populations and informal sectors.