Money laundering and the privacy design of central bank digital currency

This paper studies the implications of money laundering for the optimal design of central bank digital currency (CBDC). I build a general equilibrium framework to explicitly allow money laundering by agents and income audits by a government. I find that as long as CBDC offers less anonymity than cash, introducing CBDC will decrease money laundering. However, if CBDC still offers a relatively high level of anonymity but a low interest rate, then introducing CBDC will decrease the output from not only agents who launder money but also agents who do not. If CBDC instead offers low anonymity and a high interest rate, then introducing CBDC can increase aggregate welfare without lowering output. Furthermore, introducing CBDC needs not increase the funding costs of banks or decrease bank lending.