A Choice-Based Diffusion Model for Multi-Generation and Multi-Country Data

Abstract

We propose a framework covering multi-country and multi-generation diffusion processes. Our model focuses on choice-based diffusion model and we extend the model to incorporate heterogeneity of country. We decompose the choice probability for generation into two components; the first one is controlled by country-based variable and the other is affected by generation-based variable. This structure allows us to compare country specific differences. For empirical analysis, we estimate the non-linear random effect model for pooled data which contains 3G and 4G connection in 25 countries. Estimation results are better than several individual country models. We apply the correlated two random effects for 3G and 4G to the model. It shows the best performance for estimation period and forecasting period. The methodological contribution of our model is to simply connect both multi-country and multi-generation dimensions. This framework provides valuable information when determining policies for new country or new generation.

Keywords: Choice-based diffusion model, Multi-country diffusion, Multi-generation diffusion, Pooled model, Random effects