Stordahl, K. 2004. “Long-term broadband technology forecasting.” Teletronikk 4(4.2004): 13-31.
“This paper gives an overview of the relevant broadband technologies, describes their market positions and possibilities. Diffusion models are used to make long-term broadband forecasts for the Western European residential market. The forecasts are separated for the main broadband technologies…based on market share predictions for each technology. The long-term forecasts are evaluated and compared with other broadband forecasts for the Western Europe market” (13).
There is increasing difficulty to make specific, long-term forecasts of ICT related technologies because
the market has become deregulated.
In order to understand “broadband rollout”, we must have two components: we must first know the overall broadband coverage, and secondly the “take rate”, which is the, “proportion between demand and coverage in an area” (17).
“The long-term broadband forecasts are based on results from techno-economic calculations. The techno-economic calculations evaluate the ‘economic value’, i.e. expressed by net present value or pay back [sic] period of rollout of different broadband technologies. The assessments have been carried out for rollout on a national level and on specific areas like urban, suburban, rural and especially the rest market to examine the potential of the different broadband technologies. The main results from the techno-economic calculations show that it is difficult for new broadband technologies to capture significant market share in areas where cable modem and/or DSL are already deployed” (18).
“Important drivers in the broadband market are of course the applications. The evolution of applications generates continuously higher broadband penetration” (21). Other important drivers are the online content and the government supported programs promoting the technology.
“The broadband forecasts for the different technologies are modeled by beginning with the broadband penetration forecasts developed for the total broadband demand in the Western European residential market. Based on experience from the last few years, diffusion type models have proved to have the best abilities for long-term forecast modeling” (22).
There is a broadband forecast from 2000-2010 for Western Europe, with the inflection point at around 2005-6.
The author then compares broadband forecasts for 7 models.