TECHNOLOGY, ADOPTION AND USES

Since its inception during the 1960s the Internet has developed into an indispensable communications infrastructure. Within a few decades, it evolved from a multi-purpose to a general purpose network that can serve a broad range of uses (Bresnahan and Trajtenberg, 1995).

The development of the Internet from a network connecting a few computing centers to a global network carrying massive amounts of data was facilitated by continuous technological change in component technologies, including semiconductors, fixed and wireless networking technologies, and computing. Jointly these developments have resulted in a rapid decline of the costs of transporting, processing and storing digitized information and the ability to pack increasing computing power into smaller and mobile devices. Digitization of information flows has contributed to two waves of convergence, first between computing and telecommunications (‘telematics’; see Nora and Minc, 1978) and subsequently between telematics and media (‘mediamatics’; see Latzer, 1998, 2013). In parallel to the resulting integration or fusion of formerly separate communications activities, the sector has differentiated and diversified, even if information increasingly flows over an integrated network infrastructure. As traffic migrates to next-generation networks (NGNs) the legacy specialized voice, data, audio, and video networks are gradually being replaced and retired.²

The type and quality of Internet access and use show strong differentiations.

Fixed and wireless access networks have been continuously upgraded throughout their history but performance has increased exponentially since the 1980s, supporting ever higher download and upload speeds. Figure 1.1 shows (using a log scale) that the maximum capacity of access networks using the telephone network grew steadily from 9.6 Kbps in a typical dial-up link in 1988 to 1 Gbps in 2015 using advanced technologies such as G.fast. Similarly, mobile access networks could support 14.4 Kbps in the late 1980s but by 2015 maximum download speeds of up to 1 Gbps using LTE-A (long-term evolution-advanced) were achievable. Cable networks were initially designed for one-way delivery of video signals and therefore have historically provided broadband capacity. They continue to have a distinct lead over upgraded legacy telephone networks and mobile wireless networks, with download capacities that increased from the widespread 42.88 Mbps in 1996 to 10 Gbps by 2013. Even higher bandwidths can be provided by active and passive fiber optical networks (PONs), although they require considerable new investment. While the maximum download capacity of different platforms has become comparable, important differences between these technologies remain. For example, mobile access networks and cable networks are shared, dividing the available capacity among multiple users.

Note: ADSL = asymmetric digital subscriber line; DOCSIS = data over cable service interface specification; DSL = digital subscriber loop/line; HSPA = high-speed packet access; LTE-A = long-term evolution- advanced; VDSL = very high bit rate digital subscriber line.

Source: Own research.

Figure 1.1 Download capacity for DSL, cable, and mobile access platforms, 1988-2014

routes. With the continued migration to all-IP technology, specific network transporta­tion services will increasingly be software-defined so that they can be configured flexibly.

These performance increases and the associated price decreases have accelerated the diffusion of Internet access since the 1990s although considerable regional differences remain (Figure 1.2). The introduction of smartphones, essentially mobile computing devices, has boosted wireless and mobile broadband³ Internet use on a global scale. By 2015, there were 47.2 mobile broadband subscriptions compared to 10.8 fixed subscrip­tions per 100 inhabitants worldwide (ITU, 2015). A growing body of research takes advantage of this variation to better understand the drivers and impediments of Internet adoption and use. Although several factors, such as the intensity of competition, income and Internet literacy, are frequently identified as key drivers of Internet adoption, there is considerable variation between countries even within comparable income groups (Cambini and Jiang, 2009; Gruber and Koutroumpis, 2013; Bauer et al., 2014; Briglauer, 2014). Figures 1.3 and 1.4 show the variation in fixed and mobile broadband adoption in the OECD member states. At the end of 2014, fixed broadband adoption rates varied

Source: The World Bank, World Development Indicators, accessed 18 August 2015 at http://data.

Figure 1.2 Internet users per 100 inhabitants, 1990-2014

from 10.7 percent in Mexico to 48.9 percent in Switzerland. Mobile subscriptions, which in contrast to fixed are typically bought on an individual and not household basis, varied from 34.4 percent in Hungary to 138 percent in Finland. More recent research has gen­erated evidence that technological, economic, socio-demographic and policy factors work as constellations and that several combinations exist that may result in similar out­comes if they are appropriate to the specific national context (Groenewegen et al., 2009; Kunneke et al., 2010).

This diversity of network access platforms is also visible in the quality of the Internet infrastructure. Applications and services are becoming more differentiated with hetero­geneous bandwidth and quality needs. At the same time, network access platforms are increasingly capable of supporting high bandwidth and different service qualities, which can be software defined and software configured. One important attribute of network quality is download speed, but other characteristics such as jitter and latency also matter for certain types of applications. Figure 1.5 provides a snapshot of actually measured download speeds in OECD countries. As measurements can be taken at multiple points in the network and using different metrics, numbers generated by different providers vary. Download speeds measured by Ookla are higher than data from Google M-Labs and Akamai. The latter two are fairly consistent, as is the pattern across countries.

Enabled by more bandwidth and more widely available connectivity, the range of network uses has changed significantly over time. During recent years, real-time

Johannes M. Bauer and Michael Latzer - 9780857939845

Downloaded from EIgarOnIine at 04/05/2017 01:09:57AM via Hong kong University of Science and Technology

Notes: The OECD adjusted its definitions of fixed and mobile broadband by transferring the categories Satellite and Fixed Wireless from Mobile to Fixed Broadband.

Sources: OECD, Broadband Portal, accessed 18 August 2015 at www.oecd.org/sti/broadband/oecdbroadbandportal.htm. Information on data for Israel: http:// oe.cd/israel-di sclaimer.

Figure 1.3 Fixed broadband subscriptions per 100 inhabitants in the OECD, by technology (December 2014)

Johannes M. Bauer and Michael Latzer - 9780857939845

Downloaded from EIgarOnIine at 04/05/2017 01:09:57AM via Hong kong University of Science and Technology

150

Note: The OECD adjusted its definitions of fixed and mobile broadband by transferring the categories Satellite and Fixed Wireless from Mobile to Fixed Broadband. Mexico: Data for 2014 are preliminary; Israel, Switzerland and United States: Data for 2014 are estimates; Colombia and Latvia are in the process of accession to the OECD.

Sources: OECD, Broadband Portal, accessed 18 August 2015 at www.oecd.org/sti/broadband/oecdbroadbandportal.htm. Information on data for Israel: http:// oe.cd/israel-disclaimer.

Figure 1.4 Mobile broadband subscriptions per 100 inhabitants in the OECD, by technology (December 2014)

Johannes M. Bauer and Michael Latzer - 9780857939845

Downloaded from EIgarOnIine at 04/05/2017 01:09:57AM via Hong kong University of Science and Technology

Figure 1.5 Alternative measures of actual download speeds (Mbit/s, 1Q2O14)

entertainment (e.g., video streaming, online gaming) has been growing rapidly.

Near ubiquitous Internet access in many parts of the world has allowed migrating processing and storage of information from local machines to servers on the network (the ‘cloud’) and made it possible to develop innovative new information architectures. The Internet’s unique characteristics, including flexibility, plasticity, and scalability, have not only changed the technological basis of communications but have also had wide repercussions on all work and private activities relying on it. These features, combined with the ease of imitation in some areas (e.g., apps), have greatly intensified the competi­tive pressure in the digital economy. As many Internet-based innovations are software based or have a strong software component, the technology of innovating in the digital economy has also changed, allowing continuous experimentation via the combination and recombination of features, real-time feedback, and the rapid upscaling of successful solutions (Brynjolfsson and McAfee, 2014). Many of these effects seem to be contingent on a threshold level of adoption. Roller and Waverman (2001) found evidence that the productivity and growth effects of telecommunication networks were particularly strong in countries where a critical mass of connectivity had been reached. Koutroumpis (2009) found similar critical mass effects for broadband, although these were achieved at a much lower adoption rate.

These technological transformations have changed the economics of the Internet and have had far-reaching economic consequences for private and public sector users. Through the mid-1990s, the Internet was largely funded by the public sector. This allowed investments and technological decisions to be made with a broader public interest per­spective in mind. Once infrastructure investment decisions were predominantly made by commercial investors, this societal calculus was superseded by private profitability considerations. An overarching assumption of stakeholders was that in the competitive Internet market coordination would assure that these decisions were aligned with the public interest. There is considerable evidence in support of this premise but, like other information and communication technologies (ICTs), the Internet has a unique cost structure that interacts with the dynamics of decentralized decision-making and influ­ences the outcomes of unfettered market coordination.

With the exception of dedicated local access links, the costs of networks are shared. High upfront investment costs go hand in hand with very low incremental costs of transporting, processing and storing information. To recoup investment costs, network operators need to find ways to significantly mark up incremental costs. Pursuing such a strategy is complicated by the fact that the implementation of Internet technology has contributed to a commodification of network services. At the same time, value generation moved to the higher layers of the stack where applications and services are configured. Among the possible responses of private companies to these economic structures are horizontal and vertical integration as well as attempts to introduce differentiated network services, all major trends visible in the Internet today. The chapters in this Handbook address these challenges in more detail.

1.3