21st European Regional ITS Conference, Copenhagen 2010
This paper demonstrates how stated-preference methods can be applied to modeling consumers' preferences in the field of mobile telecommunications, and to measuring and the valuation of network effects. We illustrate this with a case study of mobile phone operators in Poland. We utilize the Choice Experiment method and present the respondents with hypothetical choices of mobile phone operators, while explicitly controlling for network effects in the form of other users in the same network. Based on the hypothetical choices consumers make we construct a conditional random parameters multinomial logit model to analyze their preferences. This approach allows us to calculate welfare effects associated with alternatives, as well as marginal rates of substitution (and hence implicit prices) of the attributes used to describe the choices, such as operator brand and distribution of family and friends between available mobile networks. The latter constitutes a network effect as consumer's utility is influenced by the number (or ratio) of members of his or her family, friends and other users subscribed to the same operator. Our results confirm the existence of a strong network effect, which is related to the size of the social network group a particular subscriber belongs to, rather than the absolute size of the mobile operator's customer base. We observe that there are two sources of this 'gross' network effect - pecuniary (arising from possible price discounts for on-net calls) and non-pecuniary, and demonstrate a way to disaggregate them. In addition, we find that brand perception and brand loyalty are important determinants of operator choice. Finally, through the application of a non-market valuation method we are able to calculate monetary values of the network effect and brand loyalty, and both turn out to be relatively high. The results might be of a particular interest to mobile phone operators and regulatory authorities - we find that the capacity for vigorous price competition between mobile operators is limited due to significant non-price barriers which mitigate subscribers' mobility in the market. We demonstrate a way to measure these effects in monetary terms based on modeling of consumer preferences.
Network effects mobile telecommunications brand valuation stated preference methods non-market valuation methods choice experiment multinomial conditional logit model preference heterogeneity random parameters model