Fixed-to-Mobile Substitution: Effects of Mobile Broadband Subscription on Fixed Broadband Termination

Motivation, background and problem statement: Fixed and mobile broadband substitutability has recently been a debate in the telecom industry as the issue affects infrastructure investment decisions of service providers and service obligation regulation decision of telecom authorities. Previous studies have debated over both the definition of substitutability, measurement and the conclusions. Theoretically, substitution is the demand side concept measured using the cross-price elasticity. With a unique dataset, this study takes a simpler approach to examine the fixed-to-mobile substitution. Instead of examining the problem through the estimation of cross-price elasticity, this study estimates the impact of users’ mobile broadband subscription on their decision to terminate fixed broadband subscription. Data and methodology: The data used in this study are from the 2016 Telecom Consumption Survey of Thailand by the National Broadcasting and Telecommunications Commission of Thailand. This study estimates the fixed to mobile substitution using the concept of average treatment effect of mobile broadband usage on the fixed broadband termination. Without random assignment, the estimation of factors determining the fixed broadband termination decision, focusing on the mobile broadband subscription, faces the endogeneity problem. Therefore, we applied the endogenous switching Probit model to estimate the average treatment effect. Results and concluding remarks: From the survey, 1949 respondents subscribed to a fixed broadband service at home at one point. Among them, 85.48 percent remain subscribed to the service and 14.52 percent has canceled the service prior to 2016. The regression analysis shows that mobile broadband subscription has a positive significant effect on the decision to cancel fixed broadband service. The contribution of the study lies on the heterogeneity of the level of fixed to mobile substitution across different groups of internet users. For those who are online for less than 40 hours per day, mobile broadband is considered as a substitute for fixed broadband to a certain level. For those whose lifestyle requires a higher usage of internet, the results show no substitution. As telecom regulations normally target to help users with lower telecom accessibility rather than those with higher level of usage, the higher level of fixed and mobile substitution for individuals with low usage should imply the possibility to unify regulatory framework for fixed and mobile broadband markets.