개발도상국의 광대역 이동통신 서비스 수요분석

초록

This study estimates the subscription demand for MBB (mobile broadband) service using a fixed effect panel data model in developing countries whose per capita GDP are less than USD 10,000. Since the quantity and the price of the MBB demand are observed only at equilibria, the 2SLS (2 stage least squares) method is used instead of the OLS method. The instrumental variables are specified as the urbanization rate or the LTE network coverage rate. It turns out that the estimation results are not significant when the time specific effect is assumed in the MBB subscription demand. Under the assumption of the country specific effect, the demand is shown to be inelastic to the service tariff and to the income at 5% significance level. In addition, except for the case where only the LTE network coverage rate was used as an instrument, the wired broadband Internet service is shown to substitute the subscription demand for MBB service at the 10% significance level. The demand elasticity to the terminal price is not obtained because the ITU data lacks the information on the terminal price for the MBB service. Even though the estimation results are limited to the developing countries, the instrumental variables method of this paper is applicable to any countries with different income levels. Of course, the instruments for these countries may be chosen differently. In addition to the endogeneity bias problem considered in this paper, it is still possible that the 2SLS method may not be suitable since many countries do not report the MBB service features to ITU. Developing an estimation method in the face of both the endogeneity and truncation problems remains a future research.

키워드