The Preference of Cash for Transactions Among Mobile Money Users in Kenya: A Machine Learning Analysis

Abstract

Digital finance platforms like Mobile Money have been a game changer in many African economies, gaining widespread uptake and improving the narrative of financial inclusion. Research shows that in Kenya, where the Mobile Money market is the largest in Africa, hundreds of thousands of lives have been improved through access to the service. At the same time, there exists evidence suggesting a rather broad preference of cash for retail transactions. Using open data from a field survey conducted in 2019, I empirically test the level of cash preference among Mobile Money users in Kenya and predict same with 11 user experience variables on 818 users in a simple classification model. Results indicate that cash preference is invariably high even among Mobile Money users, and that it is associated with negative user experiences. I recommend that the Kenyan government and service providers work together to improve infrastructural security, among other factors, to mitigate the rate of negative user sentiments that could lead to a decline in the usage of the service.