The Demand for Currency Versus Debitable Accounts : A Reconsideration

Payment choice models based on transaction sizes (TS models) induce strong predictions about the use of payment instruments. Especially, all equal-sized transactions should be paid with the same payment instrument. Then, for each individual, one should observe strict domains of transaction for every payment instruments. Using micro-level payment data from a representative sample of the French population, we show that TS models are bad at replicating individual and aggregate payment patterns. First, we show that the predictions of TS models are not empirically validated on an individual level. Second, we develop and test three models to explain the observed aggregate payment patterns. The first two models are aggregate versions of TS models and the third one is an alternative model based on a payment decision rule depending on cash holding (CH model). We find that the third model gives predictions between 2 and 6 times more precise than the first two models with notably less demanding information on individuals