Estimation of IP Telephony Demand Using the Integrated Choice and Latent Variables Approach

In Japan, demands for Broadband Internet Access and Internet Telephony have increased dramatically in recent years. According to official sources, as of September 2005, there were 21.4 million users of Broadband Internet Access and 9.76 million IP telephones in use. In this study, we employ a new generation of discrete choice models known as Integrated Choice and Latent Variable to estimate the demand for IP Telephony. The models are estimated using an original market research survey conducted in Japan during 2004 with stated preferences choice experiments of both Internet Access and IP Telephony. The Integrated Choice and Latent Variables framework allows one to include psychometric data to explicitly model the influence of attitudes and perceptions (the latent variables) on choice behavior. It leads to improved specification of discrete choice models with greater explanatory power. The methodology requires the estimation of an integrated multiequation model composed of a discrete choice model and of structural and measurement equations that permits to link the latent variables to observed indicators and explanatory variables. The integrated model is estimated using a full information maximum likelihood approach, in which the likelihood function involves multi-dimensional integrals