A Journal Article Review for "The interaction between technology, business environment, society, and regulation in ICT industries".

1. Write the Title that reflects the main focus of your work. ...

2. Cite the Article. ...

3. Article Identification. ...

4. Introduction. ...

5. Summarize the Article. ...

6. Critique It. ...

7. Draw a Conclusion.

The interaction between technology, business environment, society, and regulation in ICT industries.

Abstract

This paper focusses on the interaction between technology, the business environment, regulation, and society in ICT industries. The role of technological advances in communication (e.g., cellular mobile, 5G, spectrum allocation) and in computational advances (e.g., cloud, Internet of Things, artificial intelligence) along with developments in the business environment (e.g., disruption, convergence, Industry 4.0) and the regulatory environment (e.g., competition law and market regulation) in the model is explained. The economics of network industries and competition law and strategies such as vertical integration, bundling, and tying are described. The role of regulation and innovation is discussed along with some cases.

Introduction

A list of the most prominent businesses today, either in terms of popularity or value, would include Alphabet (Google), Amazon, Facebook, Apple, Microsoft (collectively known as GAFAM), Alibaba, and Tencent, all of which belong to the information and communication technology industry (ICT). One could also add companies such as Uber, Airbnb, Snapchat, Instagram, Pinterest, WeWork, Twitter, and WhatsApp that have become a part of people's lives.

Another common feature is that most of the GAFAM companies have raised concerns among citizens and governments about their modes of conducting business.¹In their defence, these firms claim that technological changes and new business models have made old ways of doing business defunct. The application of laws derived from the "old economy" is not suitable for the dynamic "new economy", it is argued (Teece, 2012). In the business world, there are examples of firms, such as Uber, which are yet to earn any profit, but are highly valued in the stock market (Sherman, 2019). A mythical animal, the unicorn, strides the business landscape - one that is running at typically huge losses, but nevertheless has a valuation of over a billion dollars. The ICT communication space resounds with terms such as disruption, convergence, 5G, Internet of Things (IOT), Industry 4.0, cloud and edge computing, big data, platforms, dominance, predatory pricing, standard essential patents (SEP), fair, reasonable, and non-discriminatory (FRAND), and algorithms that the common user often struggles to make sense of. Videos go "viral" and seemingly innocuous comments on social media attract the attention of trolls. As citizens, we are concerned with privacy, hacking, and fear of being exploited by "big tech companies".

There is an intricate relationship between technological developments and their impact on the world of business and society in general. Often the technological developments come with great promise, only for society to discover that the promises are accompanied by pitfalls. Then there is a tendency to try and correct the problems: either through the regulatory process or through legislation. There is always the spectre of over correction which urges caution but fails to satisfy the critics.

In this paper, we aim for a better understanding of the relationship between the technological developments, business conditions and strategies, and the regulatory environment in ICT industries and society. Changes in technology and innovation give rise to new products and services which impact the society in general and which then need to be regulated, either by regulatory agencies or the government. However, this is not straightforward and there is typically a lot of interaction and negotiation between businesses and regulators as they try to grapple with the disruption caused by innovation.

To appreciate the process, an understanding of concepts such as network effects, platforms, competition law, and vertical integration is required. Such understanding is important for society to reap the benefits of technological developments and not stymie innovative new products and processes through incomprehension and a fear of disruption.

As the very name suggests, ICT products and markets reside at the junction of information technology (IT) and communication technology (CT). The Organisation for Economic Cooperation and Development (OECD) says that ICT products "must primarily be intended to fulfil or enable the function of information processing and communication by electronic means, including transmission and display" (OECD, 2009, p11) to qualify as an ICT industry. More simply, one can say that ICT "is generally accepted to mean all devices, networking components, applications and systems that combined allow people and organizations to interact in the digital world."²

The paper is organised as follows. In the second section, we discuss the business environment in ICT industries. In the third section we present our basic model of relationships between technology, business, regulation, and society. Technological developments are described in the fourth section. In the fifth section, we discuss the economics of ICT industries, in particular, network effects and platforms. These phenomena along with the business practices described in the seventh section combine to produce effects discussed in the second section, "The business environment". In the sixth section, we discuss the regulatory environment concentrating on market regulation, competition law, the relationship between innovation and competition law, and big data and competition. This is the area where most of the current discussions regarding the relationship reside. Business strategies are discussed in the seventh section and the final section provides the conclusion.

The business environment

Disruption

Disruption is a word most associated with ICT industries (Christensen, Raynor & McDonald, 2015). In the business world, many traditional businesses are being "disrupted". Newspapers are losing ground to news portals on the Internet. With the advent of Netflix and Amazon Prime, more and more consumers are watching movies and TV shows on their computers and smartphones. This development promises deep changes in the TV network industry. A "gig economy" has emerged populated by part-time and irregular workers who provide a host of services. Changes are afoot in health care, education, banking, transportation, and a host of other industries.

Convergence

Convergence relates to the blurring of boundaries between different goods and services. In ICT industries, this is taking place at the cloud, pipe, and device level (Huang, Guo, Xie & Wu, 2012). Pipe convergence refers to the integration of Internet protocol with optical networks. Currently, networks are a hotchpotch of different technologies making them difficult to manage. Integration would lead to a number of benefits (Singh & Raja, 2010). Device convergence refers to convergence of television sets, personal computers, laptops, tablets, and mobile phones. Increasingly, devices are becoming interchangeable.

The convergence thus described is resulting in other forms of convergence. The first is service convergence. It is now possible for the same firm to combine different services into one bundle, such as telecommunications with television. A different type of convergence is that between services and manufacturing. Till now manufacturing and services have been seen as two distinct modes of production. True, manufacturing has often been bundled,³as in the case of an automobile, which requires both insurance and maintenance. However, the process is becoming more pronounced with more services such as IT and software services, design, and logistics entering into the production stage and others such as distribution and marketing at the output stage. The result is servicification of manufacturing (Lanz & Maurer, 2015).

There is one more convergence taking place in the business world. Businesses that were distinct are becoming the same. If the television is the same as a phone, then why should there be separate television and phone companies? As products converge so should businesses. The same is true for services and manufacturing. If banking moves entirely online, would banks become IT companies? It is difficult to predict the effect. Firms are entering each other's domains and entirely new businesses have emerged and are continuing to do so. There is a greater urge to merge or to acquire. Firms have to develop new strategies to succeed in the new environment (Gupta K., 2018).

The "new economy"

The term "new economy" has been in vogue in the recent past, though just what it stands for is unclear. AsL'Hoest (2001)notes, the terms "new economy", "digital economy", "internet economy", and "information economy" have been used interchangeably. Currently, the most popular seems to be "digital economy". We will continue with "new economy" since we feel that "digital economy" is somewhat restrictive. Several authors have provided characteristics of the differences between the two (Evans & Schmalensee, 2002;Posner, 2000).Table 1is a typical example.

Table 1.Differences between the old economy and the new economy.

Conventional concept	Next-generation concept
Static competition	Dynamic competition
The West and the rest	A semi-globalised world
Industry-level analysis	Ecosystem-level analysis
Vertical integration	Modularisation
Transaction and agency costs	Firm-level capabilities
Single-invention innovation model	Multi-invention innovation model

Source:Teece (2012).

One can quibble about what is included and what is not on both sides but industries such as computer software and hardware, e-commerce, mobile telephony, and pharmaceuticals characterise the new economy. In contrast industries such as energy, steel, machinery, minerals, agriculture, and chemicals are commonly thought to belong to the old economy (Evans & Schmalensee, 2002;Lazonick, 2007).

In the traditional industrial economy, most of the emphasis in competition was on price, volume, and quality. The emphasis was more in terms of competition within the market. Issues of innovation and the dynamic nature of competition were also important but usually played a smaller role. The new economy is more reliant on services; it is characterised by the use of ICT and is driven more by innovation and the introduction of new products. Dynamic aspects of competition such as innovation, research and development, and product differentiation play a more central role than competition in terms of price and market share.

Industry 4.0

The industrial revolution that began in Britain in the 19th century brought about profound changes in society and in manufacturing, not only in Europe but all over the world. This began with the invention of the steam engine which enabled mechanisation of the textile industry and division of labour, which soon spread to other industries and resulted in higher productivity. It may be called Industry 1.0. The next step, Industry 2.0, was the development of mass manufacturing using assembly lines, which began in the US as a result of consolidation of small regional markets into national markets through the development of railroads, roads, and other forms of connectivity. This also gave rise to the development of the discipline of management. Industry 3.0 centred around five inventions, the mainframe computer, the minicomputer, the personal computer, the Internet, and the smartphone. This also had and is continuing to have a profound effect on society and industry. Industry 4.0 is supposed to affect manufacturing, much of which is still in the Industry 2.0 stage. This will harness the increase in connectivity through technologies such as 5G, combined with developments such as IOT, artificial intelligence, machine learning (ML), and 3D printing. We will provide a short explanation of these technologies and then discuss them more fully later.⁴IOT refers to the development of sensors that can communicate with each other and to other computers through the Internet. This allows technologies such as self-driving cars. Artificial intelligence, as the name suggests, deals with the effort to get machines to think like humans. Machine learning is a form of artificial intelligence that designs machines to learn by themselves and 3D printing is the process of making products by adding successive layers based on a digital model. Commentators predict profound changes in society with several professions becoming redundant as a result of these developments (World Bank, 2019).

A model of interaction

The relationship between technology, business, and regulation and its impact on society is shown inFigure 1.Changes in technology are driving new forms of businesses to develop, which are in turn being regulated by the government, either directly through regulatory agencies or through laws. The interaction between the three is in turn affecting society (World Bank, 2019).

1. Download :Download high-res image (228KB)
2. Download :Download full-size image

Figure 1.Relationship between technology, business, and regulation and its impact on society.

Nor is the interaction one way. The impact on society at large from the new developments leads to pressures on governments and regulatory agencies to act to assuage problems. This leads to pressure on businesses, either from elected representatives or competition authorities. Firms react to these pressures often by vigorously defending their practices or by tweaking their products. Google's latest run-in with the Department of Justice (DOJ) in the US along with 11 other states is illustrative (Google DOJ lawsuit, 2020). The complaint notes "Two decades ago, Google became the darling of Silicon Valley as a scrappy start-up with an innovative way to search the emerging internet. That Google is long gone."⁵Google is now a behemoth, being the most prominent Internet search provider. This allows it to retain a very large presence in online advertising. Governments and regulators have been worried about the prominence of Google as can be seen from the previous actions of European Union (EU) and Indian competition authorities as well as the recent questioning of the GAFAM top executives by the US legislature.⁶

Various facets of the relationships portrayed inFigure 1have been discussed by different authors. The relationship between technology and law has till now been discussed primarily through the lens of intellectual property laws (Gifford, 2007). Of late, the relationship between competition law and innovation has been under focus (Kerber, 2017).Lianos (2019)stresses the need to think in terms of a digital eco-system to redesign competition law during the "fourth industrial revolution". The impact on society from the formation of a "gig economy" has been discussed inLee et al. (2018),Schwellnus, Geva, Pak and Veiel (2019).Lianos (2019), andCammerts and Mansell (2020)are relevant to this paper but do not consider the overall structure of interaction.

It is also interesting to note that there are interrelationships inside the boxes of technology, business, and regulation. For example, smartphones, mobile telephony, and spectrum form a mini eco-system. It is not possible to provide services at a reasonable price without a combination of these three. What complicates the matter is that the spectrum is owned and allocated by the government. Similar issues arise within IOT ecosystems (Kerber, 2019, a,b). We will dwell further on these issues after we have discussed the relevant concepts.

Technological developments

In this section, we will discuss technological and market developments with an emphasis on the Indian experience. The discussion will be divided into two sections, the first on telecommunications and the second on computing. The division we have in mind is that between infrastructure and applications that run on that infrastructure.

Communication

Mobile telecommunications

After more than two decades spent riding the voice wave, Indian cellular mobile operators have now turned to data. The overall tele-density was 84.90%, with an urban tele-density of 134.44% and a rural tele-density of 58.85% by December 2020. The rise in broadband users has been moderate, at 0.72%.⁷However, broadband usage has been increasing. Thus, telecom operators are betting on higher data usage to compensate for the decline in revenues from voice.⁸Another notable feature is that 97% of broadband connections are on mobile. So clearly a viable cellular mobile industry is vital to the development of the ICT industry.

It is difficult to overestimate the importance of cellular mobile telecommunications (or communications as a whole) for the emergence of the new economy. Since the digital economy resides on the Internet and most Indians access the Internet through their phones, a large chunk of the digital economy such as online commerce would not exist without a strong mobile industry. It has spawned a large number of businesses and has a large impact on the economy. Needless to say, it has had a very large impact on society in terms of meeting communication needs. In relation toFigure 1, we have given communication a prominent place within the technology box, which it deserves. It exemplifies the many-sided relationships, we assert, between technology, business, regulation, and government. The backward linkage between society and technology shows up in the development of apps and services such as WhatsApp. These developments in turn affect businesses and further regulation.

5G

5G affords a paradigmatic shift in the economic landscape, changing the way we produce goods to the way we conduct our lives. It will open the doors to Industry 4.0, smart cities and self-driven cars, and many other developments. Achieving that will require mobile networks of very high speeds and very low latency (the delay between an action and its implementation). For example, a self-driven car should be able to recognise obstacles almost instantly. So, the signals it receives from its sensors must be processed instantly and the instructions passed on to the controls. The situation would be similar for the provision of remote health care.

India plans to start 5G services in 2022, though Airtel says that its network is ready for 5G services and Jio expects to start services in 2021. Auctions for 5G spectrum are going to be held sometime in the future. Another issue of importance, for consumers, is the availability of smartphones that are compatible with 5G and are affordable.

The specifications for 5G are being worked on by the 3rd Generation Partnership Project (3GPP), which is a partnership of seven standard development organisations. India is represented by the Telecommunications Standards Development Society, India (TSDSI). Among the specifications are speeds of 1 to 10Gbps, one millisecond (ms) latency, complete coverage, 99.999% availability, and up to 10 years of battery life for low-power machine-type devices. Achieving these will be a challenge, given the investment required for building infrastructure and the lack of spectrum (Business Standard, 2019).

The linkages envisioned in Figure 1 are likely to get even more intense with 5G as communication extends to IOT devices. If 5G phones turn out to be unaffordable to large sections of the population or 5G networks are developed primarily in urban areas, it will exacerbate the digital divide affecting society at large. This is likely to result in government action to ameliorate the problem, further extending the interaction.

Spectrum

Spectrum refers to the electromagnetic spectrum that is used to propagate radio waves (Prasad & Sridhar, 2014). These can range from a low frequency of 10kHz to a high of 100GHz, though ultra high-frequency spectrum (UHF) in the range 400MHz to 4GHz works best. FM/AM radio, direct to home (DTH) TV and our mobile phones use spectrum to communicate. In India, we use the 700, 800, and 900MHz bands and the 1800, 2100, 2300, and 2500MHz bands for mobile services. The 3300-3600MHz band has also been recommended for auction. 5G services can be provided on the spectrum that has already been allocated but to deliver the speed and availability more spectrum is necessary. On top of that if we wish to use 5G on IOT devices then a variety of spectrum is needed. For wide area and indoor coverage for IOT devices, low-frequency spectrum of less than 1GHz is required. For standard applications that require high speed and coverage spectrum of 1-6GHz is needed. For really high speeds, one can use spectrum above 6GHz. Thus, telecom operators need spectrum in different bands to provide good services. Also, they need substantial amounts of spectrum within the bands.

Spectrum represents the second vital element in the mobile telecommunications industry. Its allocation is controlled by the government and so directly affects businesses and society. At the same time, developments in compression technology allow better use of available spectrum, suggesting a link between business, government, and technology.

Smartphones, standards, SEP, and FRAND

It is also important to note the necessity for smartphones to be able to access the Internet. Inexpensive smartphones would go a long way towards enhancing Internet connectivity. Part of the problem is that SEPs are owned by companies such as Ericsson. Building a smartphone requires adherence to standards, "a set of specifications to which all elements of products, processes, formats, or procedures under its jurisdiction must confirm." (Tassey, 2000, p 588). Standards can be adopted voluntarily or set by standard setting organisations (SSO). A standard essential patent is an invention that must be used to produce products that conform to standards.⁹Since these technologies are protected under the Intellectual Property (IP) Act, users have to pay royalties to the holders of patents. Consequently, the need for licensing on FRAND terms. Indian firms claim that they are unfairly charged for patents by firms such as Ericsson who hold the IP. In 2013 Ericsson sued Micromax for patent infringement claiming Rs. 100 crores in compensation. Micromax retaliated by complaining to the Competition Commission of India (CCI) that Ericsson was abusing its dominance. The issue raised interesting questions about the primacy of competition law over IP law (see Gupta, S, 2018). The issue was resolved through an agreement that saw Micromax paying royalties to Ericsson. These issues are likely to be compounded when it comes to standards for IOT devices.

The issue of smartphones and SEPs is another example of the complicated relationship presented earlier. In our model, we have included IP laws as a part of the regulatory structure. It is interesting to note the direction of interaction in this case. New technologies can be protected by IP laws which directly impinge on the incentives for technology development, the first box inFigure 1. This in turn affects businesses and society. Sometimes, different parts of the regulatory system, IP laws and competition authorities in this case, interact between themselves to affect the final outcome.

Advances in computation

Cloud computing

Among the factors that have helped in the transformation of the business landscape has been the emergence of cloud computing. The OECD defines cloud computing as "a service model for computing services based on a set of computing resources that can be accessed in a flexible, elastic, on-demand way with low management effort." (OECD, 2014, p 4). Business (and individuals) would earlier need to invest in both hardware and software to be able to use IT effectively. This would typically require large fixed costs of buying expensive hardware for storage and for running software programs. Cloud computing refers to the IT structure where individual firms can gain access to their software and storage needs through the Internet, also called the cloud. Essentially, it turns many of the fixed costs into variable costs.

There are some characteristics that are shared by cloud services. First is that users can avail themselves of cloud services as and when required, without human interaction and there is theoretically, limitless availability. So, firms can scale up very rapidly. Another factor that is useful is elasticity, the ability to scale down as well as up. Firms can quickly change their business models and their requirements can be met on the cloud. This allows firms to be nimble and adaptable. Third, there are no large fixed costs which have to be paid before starting to use cloud services and state-of-the-art applications are available for all types of firms. It is also possible to pay for short-term use so that fixed costs become variable costs. Fourth, since services are offered on the Internet, access can be through desktops, laptops, and phones.

A version of cloud computing, called mobile cloud computing was developed to assist owners of smartphones to access applications that would be too bulky to store on the phone. We now have access to programs such as Word and Excel on our phones, without installing the whole program. For IOT devices, versions such as fog computing or mobile edge computing have been developed. The reasoning is similar. Installing computer programs and hardware in IOT devices would make these too bulky. Better to shift the computation to the cloud. Not too deep in the cloud, because that would require greater communication, but at the "edge", possibly through virtual networks.

Internet of things (IOT)

Perhaps the simplest description of IOT is communication between machines (M2M). A simple example could be a fire detector which is connected to a computer in the fire department which alerts the concerned personnel that a fire has possibly broken out. There are a number of requirements for such a device. First, it must be on all the time. Secondly, the device would probably be idle most of the time but when a fire breaks out every second counts. It should also require low maintenance, which, among others, would mean a long battery life.

"Devices that are actively communicating using wired and wireless networks that are not computers in the traditional sense and are using the Internet in some form or another" is how the OECD (2015, p 10) defines IOT. It should be clear that for IOT to be successful 5G networks are essential.

Big data

A further confounding factor in this heady mix is the advent of big data. Broadly, the term refers to the three 'v's, volume, variety, velocity, and sometimes a fourth 'v' for value is added. Volume is readily understood as an indicator of size. However, it is difficult to appreciate how big data can get. Wal-Mart has 6000 stores worldwide, where it conducts about 267 million transactions per day. These transactions record different pieces of data, such as the item sold, the price, the geographical location, the time of sale, and buyer's characteristics. To store this data, Wal-Mart built a data warehouse to store 4 petabytes (or 4000 trillion bytes) of data in 2014 (Chen & Zhang, 2014).

Variety refers to the nature of data that is gathered. Facebook generates content in the form of text, audio, and videos. A related issue is that data are often unstructured as in the case of Facebook so that standard methods of analysis do not work. The speed at which these data are generated is called velocity. Mobile phones are constantly recording our locations and mobile phone companies are producing huge amounts of data. The fourth item - value - is qualitatively different. Big data does not possess any inherent value. Data have to be stored, relevant pieces of data extracted and cleaned, analysed, interpreted (often involving visualisation through innovative graphs and charts), and decisions made to extract value.Chen and Zhang (2014)provide a comprehensive survey on the challenges, techniques, and technologies of big data. Big data plays a nascent but prominent role in our model of interaction since it is very much at the forefront of technological and regulatory development.

Artificial intelligence (AI)

According to OECD's definition "An AI system is a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations, or decisions influencing real or virtual environments." (OECD, 2019, p 7). The raw input is data from sensors or from other sources, and the prediction is based on an objective decided by human beings using a model. The model can be built by statisticians, economists, or biologists using standard analytical tools such as regression techniques. Alternatively, it can rely on ML algorithms that learn with new data and change its predictions dynamically. Probably, the most familiar example of AI is Google searches. Similarly, Facebook and Amazon serve advertisements based on the data of past and present behaviour. Other examples would be credit scoring, predictive policing, and autonomous driving systems (Sridhar & Hari, 2020).

1. Download :Download high-res image (227KB)
2. Download :Download full-size image

Figure 2.Internet of Things (IOT) Ecosystem.

Note. M2M - communication between machines.

Source:OECD, 2016.

Machine learning and deep learning (DL) are probably the most familiar terms related to AI. ML refers to processes that learn from data. Initially, these used statistical models for predictions. Over time, they have moved to using neural networks and genetic algorithms for predictions, which is termed deep learning. The main difference is that the former depends on explicit models of the data, while the latter is more based on self-learning. In a way, the system builds its own model from the data through pattern recognition and some method of scoring good and bad predictions. Algorithms, which are simply a list of precise and simple operations to be carried out on a set of objects, are an important facet in the deployment of AI.

The developments discussed in this section belong to the technology silo in our model. As such, they directly affect businesses and their products and through that conduit affect regulation and society. Of late, there has been much interest in issues of privacy regarding usage of the Internet. There is also the notion that "data is the new oil": it is valuable in itself and with the help of AI is useful in marketing. It is feared that without access to data some businesses would be disadvantaged in relation to their bigger rivals. This is discussed more fully in the section, "Big data and competition". The rise of IOT could raise interesting issues in the technology, society, and regulation nexus in the future. Devices, networks, and software would need to work seamlessly for IOT to work, leading to the difficulty of pinning responsibility in the case of product failure. Though cloud computing is established, and the use of AI and big data has grown in the recent past, IOT is still in its infancy. The use of IOT will require standard setting and 5G networks, along with the use of AI requiring another mini eco-system (Kerber, 2019).

Economics of ICT Industries

Network effects

Network effects exist when consumers benefit from being members of a larger network than a smaller one. Other things remaining the same, it is thought to be preferable to live in a larger city than a smaller one. One would have access to more and greater variety of employment opportunities. The number of goods and services on offer would also be large and one would probably have more friends.

Something similar happens in the case of telecommunication and other networks. A new subscriber to a telephone service would be able to make calls to everyone else on the network, and everyone else would also be able to make calls to the subscriber. So, the subscriber is providing an external benefit to all existing subscribers by joining the network. In the case of telecommunications networks the benefits increase exponentially as the size of the network increases. The same situation holds for social media such as Facebook.

Network effects in ICT industries could be direct as well as indirect. For instance, a particular operating system for computers or mobile devices might become more popular. The increased popularity would lead to more software applications being written for that operating system. This is an example of an indirect network benefit (Cabral, 2017). The situation would be similar in a whole host of products such as video games, operating systems, applications (apps) for mobile phones, and search engines.

The fact that "success breeds further success" makes markets that feature network effects "tippy." A particular market may start with a few competitors, but after a point, when one producer has a sufficiently large number of subscribers or buyers, the size of the network effect is so high that everyone wants to migrate to this network, creating an example of tipping. This may happen even if the product is inferior to that of competitors or the price higher. In the extreme, there could be a winner-take-all situation with only one firm surviving.

Platforms

What are platforms?Spulber (2018, p 160) says that "(a) A platform has a 'location' that can be geographic, virtual, or some hybrid. (b) A platform has 'sides' consisting of buyers, sellers, and other groups. (c) A platform has 'intermediaries' such as market makers, match makers, and other firms that manage transactions. (d) A platform has 'transaction technologies' that handle purchases and sales, contracting, communication, market making, and matching. (e) A platform has 'coordination mechanisms' that provide incentives to participate and handle participation decisions by members of the sides of the market."

We will illustrate a two-sided market (Evans & Schmalensee, 2007;Rysman, 2009) using a brick-and-mortar example. Consider newspapers, newspaper readers, and advertisers. Newspapers provide content to readers and advertisement space to businesses that want to advertise. Clearly, a larger readership will increase the value of advertising space, which could be achieved in two ways, either by improving quality or by lowering prices or both. Consequently, there is an indirect network effect from the readership side of the market to the advertising side.

It might be optimal for the newspaper publisher to charge a lower price for the newspaper in order to capitalise on advertising revenues. The price may be lower than the cost of publishing the newspaper. In the extreme, the price may be zero or even negative. Note that Amazon and Flipkart are platforms and examples of two-sided markets. The larger the number of consumers who buy on these platforms the more the number of sellers who would like to sell here. Buyers do not have to pay for access, but sellers do. One can now understand why Flipkart and Amazon would resort to extravagant sales. These are tools to try and increase the size of the network on both sides of the market.

Network effects and platforms are important features of ICT industries. As such they do not have an overt role in our model of interaction. However, they do affect business strategies and often contribute to the rapid increase in the size of some businesses. Consequently, they contribute to the use of some business strategies (discussed in the section titled "The regulatory environment"), which can lead to action by competition authorities.

The regulatory environment

Broadly speaking, regulation is carried out by the administrative machinery of the government through its many ministries on the basis of legislation. For instance, the Department of Telecommunications (DOT) and the Telecom Regulatory Authority of India (TRAI) regulate on the basis of the Telecommunication Act. There are other acts such as the IP Act and the Information Technology Act (2000) that are also important. Various parts of the government are involved, and the list is likely to get larger as the forces of disruption and convergence play out. Here, we will deal with only two: market regulation and the use of competition law.

It is in the arena of the regulatory environment that interaction with business and society is most apparent. Businesses are the primary targets of regulatory intervention, much of it driven by public interest, it is hoped. However, that may not be so, as proponents of the economic theory of regulation have suggested. The kind of regulation that prevails is often driven by balancing different stakeholders with differing interests.

Market regulation

Independent regulatory authorities (IRA) are a relatively recent phenomenon in India, beginning in the 1990s, even though the Reserve Bank of India (RBI) was set up in 1934. After the liberalisation period in India in the 1990s, several new regulators were set up such as the TRAI, Security and Exchange Board of India (SEBI), the Insurance Regulatory and Development Authority of India (IRDA), and the Tariff Authority for Major Ports (TAMP). There are multiple actors within the regulatory sphere for telecommunications. The TRAI is one of them, but there is also the DOT and the CCI, and decisions by other ministries also have an impact. Thus, there exists a regulatory ecosystem in India.

Along with standard regulatory issues, an added question is how IRAs deal with convergence. In the case of e-banking, much of which is conducted on mobile networks, is it to be regulated by RBI, SEBI, or the TRAI, or all three? Coordination would be required, and new structures would be necessary. For instance, the Srikrishna Committee for a "free and fair digital economy" has recommended a Data Protection Authority.¹⁰IRAs act ex-ante, i.e., before the event occurs and are typically more involved and proactive, which is often desired in the fast-paced world of ICT. Given the concern that competition authorities have expressed about the efficacy of competition law to deal with GAFAM, some have suggested market regulation as a solution (see the sections titled "Competition law and innovation" and "Big data and competition").

Competition law

Competition (or antitrust) law is typically enforced ex-post (after the event) and is concerned with addressing harmful effects on competition due to anticompetitive agreements such as cartels, abuse of dominance, and mergers and acquisitions (Whish & Bailey, 2021). Competition authorities are at the forefront of the tussle between ICT businesses and society. This area of interaction has probably received the most scrutiny among academics and so is vital to our model of interaction. As noted earlier, competition authorities find themselves struggling to cater to the demands from society as well as businesses.

Competition law and innovation

Earlier we have remarked that since ICT industries markets tend to be tippy, it is usual for one giant firm or a few large firms to dominate. Examples would be Google, Amazon, and Facebook. Given the concern for monopolisation and its abuse in competition law this tendency raises concerns. One may wish for more robust intervention by competition law authorities, but there is the issue of innovation. This issue is currently the primary source of contention between competition authorities and ICT companies. As such, it also highlights the interactive process described in our model. Further, it also illustrates the evolution of economic and legal thought which has a bearing on the actual act of regulation. We will start by discussing some economic concepts that underpin jurisprudence in competition law.

The standard concept of efficiency that economists use is Pareto efficiency. An allocation is said to be Pareto efficient in terms of consumption if it is not possible to redistribute goods in such a way so that at least one person is made better off without harming anyone else. In the case of production, it should not be possible to redistribute resources so that it is possible for the economy to produce more of any good. The textbook economic structure of perfect competition is where there are a large number of sellers (and buyers) of homogenous goods with free entry and exit and perfect information is efficient. So, monopolies and oligopolies (a situation where there are a few firms) and monopolisation are a concern. Innovation is not discussed within this framework and so it is not possible to conclude that competitive markets will produce the right amount of innovation.

There are two strands of economic theory which discuss innovation. The first is the Schumpeterian (1942) view of economic development as a process of creative destruction. Firms compete with each other to develop new products and the one to succeed achieves a monopoly position. However, that position may be relatively short lived as another firm may produce a new innovation and in turn become the new monopolist. In Schumpeter's scheme of things competition within the market is relatively less important than the competition for the market; the process of innovation. A monopolist cannot rest on his laurels since there is always potential competition. This is an argument frequently used by technology giants to justify their overwhelming presence in some markets. They are still competing frantically against the unknown innovator who may appear out of nowhere and make them irrelevant. Schumpeter's other hypothesis also seems to bolster their arguments. He opined that larger firms are more likely to innovate. Even though Schumpeter's arguments are often invoked they have not been sufficiently developed in the economics literature.Arrow (1962)shows to the contrary that a monopolist may have a lower incentive to innovate. A competitive firm would have the incentive to become a monopolist if the innovation was sufficiently important, while a monopolist would be merely replacing himself. Consequently, competitive markets would do better in fostering innovation rather than monopolies. There are several empirical studies (Aghion, Bloom, Blundell, Griffith & Howitt, 2005;Hashmi, 2013) that say that the relationship between market concentration and innovation is complex and likely to be U-shaped. There is also a large amount of theoretical work within industrial organisation literature that looks at research and development and patent races (Bundeskartellamt, 2017;Encaoua & Hollander, 2002). However, not much of it is helpful in guiding competition law (Kerber, 2017).

The argument at the heart of the debate is between overenforcement and underenforcement of competition law. Too strict an adherence to the established doctrines of competition law may end up killing the goose that lays the golden eggs. Critics of competition law authorities argue that their methods and philosophy are outdated, as they rely on the static foundations of competition laws and that is inappropriate for the dynamic nature of the new economy (Kerber, 2017;Shelanski, 2013). This overstates the issue because even in conventional markets competition law authorities have been awake to dynamic issues. Underenforcement, may on the other hand allow monopolies to grow so large that they are essentially beyond control and require drastic action.

One view is that the forces of competition and innovation will automatically correct any short-term positions of monopoly and that competition authorities can afford to lower their guard. The opposite view is that given the inherent impetus towards monopolisation and its attendant abuses in the new economy, competition authorities should be even more vigilant. It is vigorous competition that best promotes innovation (OECD, 1996). The first view was in vogue earlier but now competition authorities seem to be leaning towards the latter. This is an example of how regulatory thinking evolves with the development of new technologies and their use in businesses.

Recently, the competition authorities of Germany, the UK, the EU, and Australia have published reports that attempt to deal with the challenges posed by the digital economy (Kerber, 2019). Among other matters, they have considered whether competition laws need to be amended or if there is need for a market regulator like the TRAI. The Government of India (Ministry of Corporate Affairs, 2019) has also recently released a report reviewing the competition laws of India, where there is a chapter on "technology and new age markets" and the CCI has published a report on e-commerce in India (Competition Commission of India, 2020).

If we were to consider the record of the CCI from its judgements and its reasoning in its orders, it is clear that it is cognizant of the regulatory costs of being overzealous. There were a number of cases of alleged abuse of dominance. Its most famous case involved Google, which was found dominant in web search and web advertising and guilty of abuse. There were cases against Flipkart of predatory pricing which were dismissed. WhatsApp was accused of abuse over privacy issues, and is currently being investigated.¹¹Accusations of exclusive dealing were made against Flipkart, MakeMyTrip, and Oyo. One was dismissed and the other is being investigated. Allegations of collusion were made against ANI Technologies/Uber for their use of algorithms in pricing, which were dismissed. Another issue that was highlighted was the hub and spoke mechanisms used by online sellers. There was also a case of resale price maintenance against Kaff, that was dismissed.

There is a complicated interaction between technological developments, their analysis by economists, action taken by regulatory authorities and the legal profession and society. It is interesting to follow the twists and turns of this process through the academic literature (Gilbert, 2006;Ginsberg & Wright, 2012;Hoffman, 2019;Newman, 2019;Ohlhausen, 2017;Posner, 2000;). AsKerber (2019)notes, a number of competition authorities now recognise the depth of the problem posed by market dominance by digital superpowers but come up with different solutions. Some scholars have advocated a new "protection of competition standard" (Shapiro, 2019;Wu, 2018). As of yet, no clear methods and doctrines have emerged.

Big data and competition

The effect of big data on competition and consequently on competition law is not clearly understood at present. Again, this represents another aspect of our model of interaction, though one that is in its infancy. The most comprehensive analysis can be found in a paper written by the French and German competition authorities (Competition Law & Data, 2016).¹²Sokol & Comerford, 2017argue that there have been no cases in Europe or the United States where it has been shown that big data actually harmed competition.Stucke and Grunes (2015)assert the opposite: that most of the claims of the harmlessness of big data are false.

The anticompetitive effects of big data could be due to network effects. Google is the most used search engine and it may be the case that such usage allows it to provide better results than other search engines. As more people search on Google, it would generate more data and better results, attracting even more to do so, Further, lack of data could hinder new search engines from being introduced, a form of barrier to entry.

Google gets its revenues from advertising, an example of a two-sided market. This leads to a strengthening of the network effect as higher usage attracts more users and consequently more advertisement and possibly higher advertising rates, accentuating dominance. Further, it could use big data to anticipate possible entrants and then either crush its nascent competitors, by denying them access to data, for example, or to acquire them.

A more recent worry is the use of algorithms (OECD, 2017). Businesses such as travel portals use dynamic pricing algorithms; a method where the price of products changes continuously according to demand and supply conditions. If businesses in the same industry use similar algorithms to set prices it raises the possibility of collusion, both explicit and tacit (where firms do not have an explicit agreement but nevertheless act in a coordinated fashion). If all firms use similar algorithms their actions are likely to mirror their rivals. Also, it may make detection and punishment easier and so facilitate collusion. There is also the possibility of personalised pricing, where prices and quantities are unique for each individual, making it difficult to detect abuse of market power.

Importantly, there is also the issue of privacy on which the views are split. One school of thought regards privacy as a form of non-price or quality competition. The opposite view is that privacy concerns should be left to consumer protection laws and competition laws will only serve to muddy a complicated issue. There are other forms of non-price competition beyond privacy. One possible example is that of manipulating search results to favour certain firms. The effect is that of lower quality of search results, which a dominant firm can get away with because of its position.

The outlook is not all bleak on the competitive effects of big data. First, big data has enabled fine tuning of products and services that are delivered to us, particularly in online markets. Many services would not have existed without the extensive use of data, such as navigational tools. Also, there are large amounts of data being generated that data is not in the control of a few firms. There are a large number of firms that sell data and prices are low. Information is like a public good in that it is non-rival. The same data can be used by multiple businesses without losing its value. Further, consumers usually use a number of service providers, a practice known as multi-homing. One may use both Google and Bing, Amazon and Flipkart, Myntra and Snapdeal. All these firms will produce similar data, even though of different volumes. Finally, big data is only one aspect of competition. Firms can introduce new products and have innovative business strategies to overcome the big data advantage of rivals.

Strategies for gaining market power

In this section, we will briefly discuss the different types of strategies that firms might employ in pursuit of monopolisation. These strategies are not unique to the new economy, except that features of the new economy make it particularly amenable to these methods. Also, these strategies are by no means illegal but that they can be used to extend market power, which can then lead to abuse.

Vertical integration

Vertical integration occurs when a firm operates at two or more stages of a production process that can be thought of as distinct. For instance, a manufacturer who sells his own products through retail stores would practice vertical integration. The effect of vertical integration on efficiency is difficult to judge. On the one hand, vertical integration eliminates the familiar double marginalisation problem. For example, a downstream retailer will buy his goods from the upstream manufacturer. The price that the manufacturer charges becomes the cost of materials for the retailer. The retailer would add a margin to this cost to arrive at the final sales price. Thus, margins are added on at two different stages and the total sales is lowered and prices are higher compared to one integrated firm. So vertical integration should be good for consumers (Cabral, 2017)

However, competition authorities have always viewed vertical integration and practices such as market foreclosure, price discrimination, and tying, that are sometimes used by vertically integrated firms,¹³with suspicion. The fear is that a firm may be able to extend its market power in one stage of the value chain to another. The current theory suggests that vertical integration can be welfare enhancing as well as the opposite and empirical evidence about damages is scarce (Owen, 2011). However, it is possible that in the context of the new economy there may be reason to worry. It is possible that convergence, network effects and the tendency to tip can make it easier for market power to be transmitted vertically, as well as horizontally (Evans, 2020).

Margin squeezing and refusal to deal

Here, we will discuss two practices that might facilitate the acquisition of market power through vertical integration: margin squeezing and refusal to deal. The first is a form of price discrimination. A vertically integrated firm may charge a higher price to its competitors in the more competitive market if they are dependent on it for the inputs in the less competitive market. An extreme form of this practice would be to refuse to deal. The possibility of such occurrence resulted in the regulatory innovation of the essential facilities doctrine, which guards against such practices by providing access to such facilities at a reasonable price.

Bundling and tying

Bundling and tying may also seem innocuous. Bundling is the practice of selling a number of different goods as a single package with a single price. The goods may be physically similar such as a package meal served in restaurant. One may be able to buy the individual items of the meal separately, but they are also available as a bundle. They may be very dissimilar, such as bundling sales of refrigerators with an air ticket. The practice of tying requires a consumer to buy essential inputs for the operation of a product from the manufacturer. The standard example would be that of Xerox requiring consumers to buy toners from it.

As before, the threat comes from the nature of new economy. Bundling could become a strategic tool and pure play operators who specialise in only one service may lose out. In some cases, some items in the bundle may be sold at zero price possibly to take advantage of network effects.

Predatory pricing

Quite frequently, firms in ICT markets are accused of predatory pricing. Predatory pricing refers to the practice of keeping prices artificially low so as to facilitate the exit or prevent the entry of competitors.

In ICT markets, we have to be even more wary of the charge. Because of network effects firms often charge low prices initially to increase adoption and the size of the network which subsequently increases the value of the product to the consumer. So initial low prices could be a legitimate business practice. As we have argued earlier much the same can be argued for two-sided markets. Charging low or even zero prices to one side of a platform should be legitimate. Consumer lock-in, where individuals find it difficult to switch suppliers because of inconvenience, is more difficult to judge. The possibility of lock-in makes consumers more valuable to firms since they can expect to milk them later. However, that possibility makes firms compete harder to get these consumers, possibly through lower prices. Forward-thinking consumers should be aware of the possibility of lock-in and bargain appropriately or keep their options open.

Conclusion

We have discussed at length the technological developments that have taken place as well as the challenges that confront competition law authorities and businesses. Here, we want to provide some points to consider as they go about their work. First, businesses have to pay heed to the concerns of governments and the society in which they operate. It is usually the case that governments and society, with some exceptions, are behind the curve. Technological developments take place very rapidly while regulatory structures and legal doctrine evolve much more slowly (Coops, 2005). It would be very difficult to find judicial officers and lawyers who are fully conversant with developments in technology and economics. The same would be true of regulatory bodies, civil servants, politicians, and the general public. In their quest to capture markets through disruptive technologies and innovative business models these businesses would do well to explain their actions and strategies to other stakeholders.

Governments, regulators, and members of the civil society have to stay abreast of the developments in technology and have some understanding of the digital economy. They need to find new innovative ways of regulation and need not be too wary of old-style regulation either. Emergence of new technologies such as IOT will add another layer of complexity as standardisation gains centre stage. Competition and regulatory authorities need to tread a cautious path as activities that were viewed as anticompetitive in the old economy, may actually be welfare enhancing in the digital era. So, the dynamic interaction between technological advances, their use in new business models, their effects on society and efforts to regulate them will continue. It is necessary to get used to it.

References

Aghion et al., 2005P.Aghion,N.Bloom,R.Blundell,R.Griffith,P.HowittCompetition and Innovation: An Inverted-U relationshipQuarterly Journal of Economics,120(2015)(2005), pp.701-728View Record in Scopus

Arrow, 1962K.J.ArrowEconomic welfare and the allocation of resources for innovationR. NelsonThe rate and direction of economic activity: Economic and social factors,Princeton University Press,Princeton, NJ(1962)

Bundeskartellamt 2017Bundeskartellamt (2017).,Innovations - Challenges for competition law practice. Series of papers on "competition and consumer protection in the digital economy",https://www.bundeskartellamt.de/SharedDocs/Publikation/EN/Schriftenreihe_Digitales_II.pdf?__blob=publicationFile&v=3

Cabral, 2017L.M.B.CabralIntroduction to industrial organizationMIT Press,Massachusetts(2017)Google ScholarCammerts and Mansell, 2020B.Cammerts,R.MansellDigital platform policy and regulation: Toward a radical democratic turnInternational Journal of Communication,14(2020), pp.135-154

Chen and Zhang, 2014P.C.L.Chen,C.ZhangData-intensive applications, challenges, techniques and technologies: A survey on big dataInformation Sciences,275(2014), pp.314-347CrossRefGoogle Scholar

Christensen et al., 2015C.M.Christensen,M.E.Raynor,R.McDonaldWhat is disruptive innovationHarvard Business Review(2015)December 2015https://hciforstartups.org/wp-content/uploads/2021/10/What-Is-Disruptive-Innovation.pdf

Competition Commission of India 2020Competition Commission of India. (2020).Market study on e-commerce in India. Key findings and observations. 08-01-2020.https://www.cci.gov.in/economics-research/market-studies/details/18/6

Competition Law and Data 2016Competition Law and Data. (2016). Autorite de la concurrence and Bundeskartellamt., 10th May 2016.https://www.bundeskartellamt.de/SharedDocs/Publikation/DE/Berichte/Big%20Data%20Papier.pdf?__blob=publicationFile&v=2

Coops, 2005M.J.CoopsDisruptive technology...disruptive regulationMichigan state law review,1,Spring(2005), pp.309-3162005Google ScholarEncaoua and Hollander, 2002D.Encaoua,A.HollanderCompetition policy and innovationOxford Review of Economic Policy,18(1)(2002), pp.63-79View Record in Scopus

Evans, 2020D.E.EvansVertical restraints in a digital world in the evolution of antitrust in the digital era: Essays on competition policyD. S. Evans, A. Fels, & C. Tucker (Eds.)Competition policy international(2020), pp.1-49https://www.nber.org/system/files/chapters/c10784/c10784.pdfView Record in Scopus

Evans and Schmalensee, 2007D.S.Evans,R.SchmalenseeThe industrial organization of markets with two-sided platformsCompetition Policy International,3(1)(2007)Spring

Evans and Schmalensee, 2002D.S.Evans,R.SchmalenseeSome economic aspects of antitrust analysis in dynamically competitive industriesInnovation Policy and the Economy, NBER,2(2002), pp.151-179CrossRef

Gifford, 2007D.J.GiffordLaw and technology: Interactions and relationships. 2007Minnesota Journal of Law, Science & Technology,8(2007), p.571Retrieved fromhttps://scholarship.law.umn.edu/faculty_articles/324Google ScholarGilbert, 2006R.J.GilbertCompetition and innovationW. D. Collins (Ed.)Issues in competition law and policy,American Bar Association,Chicago(2006)Antitrust Section

Ginsberg and Wright, 2012D.Ginsberg,J.D.WrightDynamic analysis and limits of antitrust institutionsAntitrust Law Journal,78(1)(2012), pp.12-48

Google DOJ lawsuit 2020Google DOJ lawsuit. (2020). Retrieved fromhttps://www.justice.gov/opa/press-release/file/1328941/downloadGoogle ScholarGupta, 2018aK.GuptaFRAND in India: Emerging DevelopmentsIIMB Management Review,38(1)(2018), pp.27-36March 2018ArticleDownload PDFView Record in Scopus

Gupta, 2018bS.GuptaDriving digital strategyHarvard Business University Press,Cambridge, MA(2018)Google ScholarHashmi, 2013A.R.HashmiCompetition and innovation: The inverted-U relationship revisitedThe Review of Economics and Statistics,95(5)(2013), pp.1653-1668December 2013View Record in Scopus

Hoffman, 2019D.B.HoffmanAntitrust in the digital economy: A snapshot of ftc issues - Remarks at gcr live antitrust in the digital economyDirector, Bureau of Competition, U.S. Federal Trade Commission(2019)https://www.ftc.gov/system/files/documents/public_statements/1522327/hoffman_-_gcr_live_san_francisco_2019_speech_5-22-19.pdf

Huang et al., 2012I.Huang,R.Guo,H.Xie,Z.WuThe convergence of information and communication technologies gains momentumThe Global Information Technology (Report No. 2012),INSEAD and World Economic Forum(2012)p://www3.weforum.org/docs/GITR/2012/GITR_Chapter1.2_2012.pdfLast accessed April 17, 2017Google ScholarJin and Wagman, 2019Jin, G.Z., .& Wagman, L. (2019).,Big data at the crossroads of antitrust and consumer protection.https://lwagman.org/BigDataAntitrustCP.pdf

Kerber, 2017W.KerberCompetition, innovation and competition law: Dissecting the interplayMAGKS Discussion Paper No. 42-2017(2017)Google ScholarKerber, 2019aW.KerberData-sharing in IoT ecosystems and competition law: The example of connected carsJournal of Competition Law and Economics,15(4)(2019), pp.381-426CrossRefView Record in ScopusGoogle ScholarKerber, 2019bKerber, W. (2019b). Updating competition policy for the digital economy? An analysis of recent reports in Germany, UK, EU and Australia. (Abstract).https://www.papers.ssrn.com/sol3/papers.cfm?abstract_id=3469624#

Lanz and Maurer, 2015R.Lanz,A.Mau