Technological diffusion: What can we learn from the case of worldwide mobile telephony?

The bottom of the pyramid literature advises multinational companies to launch disruptive innovation in less developed countries before launching it in developed countries. One reason for success in less developed countries is that a disruptive technology may spread faster thanks to leapfrogging; a reason for lack of success is that people with lower incomes may adopt technology slower. We examine these hypotheses by comparing the diffusion of mobile telephony in 183 developed and less developed countries. A discriminant function analysis indicates that on average the speed of mobile telephony diffusion (calculated using Griliches's epidemic diffusion model) is highest in low-income countries, then in middle-income countries, and lowest in high-income countries. We find that mobile telephony diffuses fast in the former communist European countries and in many Sub-Saharan African countries, but rather slow in North America, the Nordic region, Oceania, and East Asia. The causes for the difference in the speed of mobile telephony diffusion are similar in low-, middle-, and high-income countries. For data covering all countries our findings indicate that higher telephone mainline penetration in the year of mobile telephony introduction increases the speed of diffusion of mobile telephony; that the later mobile telephony was introduced, the faster it got adopted; and that a higher income per capita in the year of mobile telephony introduction decreases diffusion speed (contrary to expectations). However for data covering only high-income countries the relationship between variables is opposite: a higher telephone mainline penetration decreases the speed of diffusion of mobile telephony, and a higher income per capita increases it.