In India, internet advertising is gaining significance with the number of internet users growing fast. As of 2013, internet users have increased by 31% over the previous year, with the total number of users crossing 150 million (Hindu 2013). It was estimated that India was likely to be the second largest market globally in terms of number of internet users by 2016, just behind China (Financial Times 2014). With such a growth rate, it can be expected that the role of online advertising is set to be significant in India in the foreseeable future. Starting with the first clickable advertisement that came out in Hotwired in 1994, the online advertisement industry has grown to become a significant player in the media advertising industry. In 2015, global internet advertising expenditure was estimated to have reached more than $120 billion, around 25% of the total advertising spends, including mobile advertising (Lunden 2014). Advertisers use the internet to reach consumers in many ways, including display advertisements (shown on websites that a user visits) and search advertisements (where advertisements correspond to a search query the user had entered). Search advertisements are used by general search providers like Google, Yahoo, Bing, and others as well as specialty providers like Amazon (shopping) or Expedia (travel) and accounted for 45% of the overall internet advertising market in 2014 (Lunden 2014).
In this context, the Competition Commission of India’s report suggesting that Google indulged in certain anti-competitive practices, becomes relevant (Gupta 2015). Briefly, the main contentions are as follows: Google is alleged to have a dominant position in the search engine market because it is the most used search engine in India. Further, it is also alleged that Google has abused that dominant position by promoting its other products (built-in services) through its search engine such as maps, shopping, etc. Such promotion has supposedly denied competitors in these built-in services markets a chance to compete because Google’s products appear when a searcher searches for those services. In order to comprehend these allegations from an antitrust perspective, it is important to understand how firms compete in markets like these that are characterised by fast and ever-changing technology. As the United States Federal Trade Commission has pointed out, “Technology industries are notoriously fast-paced, particularly industries involving the Internet. Poor or misguided antitrust enforcement action in such industries can have detrimental and long-lasting effects” (Ohlahausen 2013).
Platforms like Google are referred to by economists as two-sided or multisided markets, which facilitate interactions between various parties. Given their nature, business models underlying such platforms tend to be different from a more traditional firm. In this context, the purpose of this paper is to elaborate on the economic arguments that arise with respect to such online platforms and raise questions that need to be answered in order to evaluate the allegations against Google in India. This understanding is of interest not just to the competition policy and law practitioners, but also to researchers in competition economics at large. While allegations against Google range from infringing of trademark and copyright laws to abusing a dominant position (Ganz 2015; Wang and Wang 2016), in this paper, I focus only on the abuse of
Economics of Two-sided Markets
As a first step in looking at the antitrust issues, it is important to look at the economic principles that govern platforms, where the operator benefits by bringing together various groups and creating economic value by their interaction. Economists refer to these as two-sided markets or multisided markets.1 Platform markets work, perhaps because, they may experience both direct and indirect network effects. Direct network effects arise when the more people use the platform, the more people on that side of the platform benefit. For example, users are more likely to visit Facebook if their friends and family also use it. By contrast, indirect network effects work when a change in dynamics of one side of the platform improves the overall welfare of the platform through incentivising the other side. Such network effects, if they exist for a given platform, alter traditional antitrust concerns so much so that the concerns might not even arise. Antitrust analysis of platforms must account for the interlinkages between the various sides that the platform connects, and traditional economics models and tools used in antitrust analysis might be applicable with significant modifications based on the context. I discuss these concerns in the subsequent sections.
Internet search is a two-sided platform: Two-sided/multi-sided markets have been in existence for a long time. Traditional media outlets like television, newspapers, and Yellow Pages are some such examples.2 For example, while the price of a newspaper may be less than the cost of printing it (cost of newsprint plus the cost of gathering news, etc) the newspaper business can still be profitable because the lower price of newspapers attracts a greater number of buyers, which in turn, generates greater revenue from advertising. Similarly, a search engine offers the ability to search for free and increases the user base in order to command greater price from the advertising market.
However, an important question is: What distinguishes search engines from the other platform markets? The first major distinction is the fact that the advertisements are more targeted in nature. That is, since the advertisement appears conditional on the usage of certain search terms, it can be presumed that the searcher has more interest in the products being advertised than, say, a random newspaper viewer. For example, the probability of me buying a cricket bat is higher when I see an advertisement for a cricket bat while I am searching for information on cricket bat, than when I am watching an advertisement of a cricket bat on television in the middle, of say, news about stock market. Further, sophisticated algorithms involving the searcher’s past online behaviour can be employed to further tailor advertisements to the searcher’s interests. Moreover, a viewer of advertisements during a search process can be directly taken to the point of sale, as against traditional media advertisement. Given these factors, search engine advertisements enjoy some key benefits when compared to other advertising.
The other major distinction between the search engines and other platform markets is that search engines are free. Developing search engine is an expensive proposition, requiring huge amounts of sunk investments. At the same time, not all consumers who use the “search/cost side” of the platform are consumers of the “revenue/advertisement” side of the platform. That is, there is a significant searcher heterogeneity that needs to be taken into account in order to analyse search engine behaviour. Further, while sunk costs are immense (maintenance of search engine), variable costs are miniscule. Both these points suggest that the larger the searcher database is, the larger is the revenue. That is, the number of searchers has to cross a certain tipping point for this to be profitable. Given these, search engines target their business strategies in making their searcher database (the users of free end of the platform) as large as possible.
As already pointed out, search engines are a two-sided platform with internet users on one side and internet advertisers on the other. Barring some operational differences, the essential source of revenue for search engines is through selling advertising space, as search engines are invariably free-to-use.3 Offering the search service to users is a cost the search engine bears. For example, on the search side, Google uses proprietary and complex algorithms in order to address queries from the searchers at no charge. On the advertising side, Google uses a product called AdWords that sells advertising space displayed along with the search results to companies based on the search string of words employed by the searcher. Advertisers bid for space on a per-click basis, that is, Google gets paid a certain sum of money whenever a user clicks on the advertisement. The pay-per-click mechanism generates some other strategic incentives for Google, than to simply look at the price bid by the advertiser. For example, it makes economic sense to place an advertisement that is more likely to be clicked upon at a lower per-click rate that an advertisement that will not be clicked upon even if per-click rate is higher. As per my understanding, the auction works as follows: for every search string employed by the user, all advertisements that match those keywords are selected. Then, using a proprietary algorithm, the advertisements are ranked (presumably on the estimate of revenue the advertisement can generate). Then, the top advertisements are selected, along with the ad positions where the advertisement would appear.4 The advertiser who is interested in the searchers that employ a given search string, bids for how much it is willing to pay for every click on the advertisement registered via the search engine.
Search engines do not operate like standard markets with direct network effects (Rochet and Tirole 2003; Evans and Schmalensee 2014). Instead, search engines exhibit indirect network effects where changes in the dynamics of one side of a platform (for example, searchers) has a significant impact on the other side (online advertisers). If the number of searchers on the platform increases, advertisers find it more advantageous to advertise with that platform (Evans 2010). This is because, the more searchers there are, the higher the probability of finding a potential consumer. This can make advertising on the platform more effective in terms of generating revenue for the seller and may thereby translate into a greater willingness to pay by the advertiser.5 However, the reverse is not necessarily always true. It is generally not the case that the more the advertisers on a platform, the more attractive the platform is to users. Given that there is negligible switching cost across various search engines (another search engine, is after all, a click away), users tend to gravitate towards those search engines that offer the best possible search content. In other words, search engines with superior search content (cost side of the platform)—as against the revenue source of advertising—tend to attract more searchers. Therefore, the main takeaway is that the precise nature and balance of these markets imply that for search engines to maximise their revenue, they have to incur substantial costs towards innovation aimed at making their search engine superior to that of their rivals.
To sum up the preceding discussion, traditional economic analyses adopted by the authorities in evaluating various concerns relating to antitrust law may not be applicable. The main objective of a search engine is to maximise the overall value of the connections between the different sides of the platform, including taking advantage of indirect network effects. To this end, search engines employ several strategies that are in contrast with the markets not characterised by this two-sided nature, including pricing significantly below marginal cost for one side. This is the reason search engines make searches free for users, even if it is costly to provide the search service. Search engines instead earn revenue by working to increase the expected return on the amount that companies pay for advertising and therefore increasing their willingness to pay to reach the search engine’s users. Any strategy that would interfere with this indirect network effect—where advertising space is made more valuable by attracting more users who are likely to click on advertisements—would be an irrational choice.
Websites in results are not market participants: It is important to understand that the websites in Google’s search results are not participants in the two-sided platform served by Google’s search engine platform. Since the primary goal of a search engine is to connect potential buyers with the potential sellers, the market participants constitute advertisers and the searchers alone, and not the websites that show up for any generic search. These search results are only a means towards achieving the end objective of connecting advertisers and searchers. In this context, Salinger and Levinson (2015) argue,
For any Web Site that wants to attract viewers, Google generates a positive externality. Web sites are free to base their business model on continuing to receive this externality, but that business decision on their part neither makes them a Google customer nor creates an obligation for Google to continue to provide that externality.
In sum, they argue that these websites are not a part of core consumer group, but yet, receive a free service (even if they did not request for such service) and it would be unfair to expect Google to accommodate their request at the cost of hurting its own interests. Understanding the various actors in the market transaction also from a policy perspective tells us what to examine for purposes of competition law harm.
Market Definition on Both Sides
Before analysing search engines from an antitrust perspective, it is pertinent to understand the relevant market. One of the tools that economists and antitrust authorities have used in order to define appropriate market for analysis is, “cross-price elasticities” in order to define appropriate market for analysis. The economic logic is that every other product/market affected by a change in the dynamics (for example, price, quality, diversity, etc) of the product/market under question ought to be considered a relevant market. This question becomes more complex in the context of two-sided markets generally. This section examines competition on both sides of the internet search engine market and considers whether a relevant market can be defined in this case.
Competition for users: Search engines compete for searchers’ attention not just among themselves, but depending on what the searcher is looking for, several other services enter into the picture. Consider, for example, a potential searcher who is looking for a vacation package to the Andaman and Nicobar Islands. The searcher has a choice between using various search engines like Google, Bing, etc, or going only to travel-related websites like MakeMyTrip, Yatra, etc, for flight deals, and Trivago.com and Hotels.com for accommodation packages. At the same time, other advisory websites like Facebook, TripAdvisor, etc, that provide information on user experience in this destination also compete for this searcher’s attention. Even more significantly, the range of competing websites and services can change significantly depending on what the user is looking for. The travel-related services listed earlier would be of no help to a searcher looking for information on purchasing, for example, a new smartphone. Instead, the searcher would choose among an entirely different set of services such as shopping or product review websites like Flipkart, Amazon, or Infibeam. It is also important to note that unlike the case of Microsoft’s dominant Windows operating system, where users faced high costs to shift to another operating system, searchers’ cost of switching from one search engine to another (and in general for switching to other sources of information like those described earlier) is minimal (Edlin and Harris 2013). The low barrier to switching further expands the range of potential sources of information that may be affected by change in the competitive dynamics of general search engines like Google. It is, therefore, not clear how to define a market for search engines in general or even if such a market really can, indeed, be defined in the first place. At a minimum, any analysis of Google’s search engine must account for the different groups of services that searchers might choose for different kinds of search queries.
Competition for advertisers: On the advertiser side too, general search engines compete with a wide range of other options available to advertisers. In the nascent stages of evolution of search engines, one could have plausibly argued that internet searches were a separate market on their own, given the targeted nature of their advertising. However, with the emergence of websites in the social media space, more sophisticated technology based on identifying personal characteristics of a website user, display advertising is becoming more targeted and a closer substitute for search engine advertising. In 2014, display advertising accounted for almost 43% of the total online advertising market.6 Continuing the same aforementioned example, a hotel in Andaman might consider advertising on websites like Facebook, Twitter, etc, or more specialised sites like Andaman’s official tourism website instead of advertising on any search engine. Any reduction in quality of searches provided by the search engine is, therefore, likely to drive searchers away from the search engine. Such reduction in the number of searchers results in alternative forms of advertising becoming relatively more valuable for advertisers. Hence, the argument that search engines belong to a different class may not be valid any longer. Researchers have argued that, since the primary reason for advertisers to pay for advertising is to “purchase sales,” several types of advertising “with varied characteristics can nevertheless compete with each other on price” (Ratliff and Rubinfeld 2010).
As with searchers, analysis of the advertiser side of the platform must account for the many different services that companies might use instead of search engine advertising. Empirical studies that look for cross-elasticities of advertising expenditure are needed in order to arrive at a consensus on this issue. To my knowledge, there is no such consensus among the academic community at this point. Competition law analysis limited only to competition among search engines for advertisers could misunderstand the dynamics of online advertising and therefore not be relevant. Indeed, it may misanalyse how competition works and may lead to mistaken inferences as to potential harm.
Evaluating Product Changes to Google Search
The primary allegation against Google is that it has leveraged its search engine platform to promote Google’s other specialised search products. Two questions arise in this context: (i) are Google Search and rest of Google products indeed separate entities? (ii) are the consumers of Google’s search engine platform harmed in anyway because Google has been providing these built-in services?
Google product improvement: Product differentiation and technological innovation through facilities to make the platform more attractive for any group of consumers is a common strategy in two-sided markets (Rysman 2009). In the search engine context, these product improvements could include features that make the experience better for either searchers or advertisers. For example, Google has employed a number of strategies to improve the advertising experience on its platform. Auctioning the advertising slots on a pay-per-click basis can be considered one such pricing strategy. Presumably, the advertiser knows more than the search engine on the impact of advertisement, and in such situations an auction is a more efficient way to price in the market. Placement of advertisement on the search interface is another strategy employed by the search engines (Evans 2010: 210–12).
With respect to searchers, the discussion on how search engines work—discussed in an earlier section—shows that search engines compete to maximise the quality of their results, in the presence of competition from other search engines and other online services because of low switching costs. One way to increase the attractiveness of the platform to searchers, and therefore to advertisers, is through incorporating changes to its product from time to time. These products are aimed at both differentiating itself from the competitors, as well as at making search experience better. As discussed in an earlier section, search engines compete with a wide variety of services—including both other search engines and more specialised services—for searchers’ attention. Therefore, one could consider built-in services that improve a search engine’s ability to respond to specialised queries as an attempt to differentiate by providing a better search experience. Researchers have argued that Google providing additional services along with its search results (calculator, maps, weather, etc) are innovative attempts by Google in this direction (Salinger and Levinson 2015). In other words, Google’s built-in specialised search features are not separate products but an improvement to Google’s search engine platform.7 Other search engines available in India like Bing employ similar strategies to a certain extent.
Searchers and advertisers benefit: A next question that needs to be considered in this context is, whether or not Google’s consumers are hurt through Google’s use of specialised search results. As discussed earlier, since any search engine is an intermediary between searchers and advertisers, the consumers of Google’s platform are searchers and advertisers: not the websites that appear in Google’s search results. A searcher would only be harmed if the search results provided by Google decrease in quality, which is counterproductive for Google for several reasons discussed earlier. The cost for a searcher to switch away from Google is very low, so a reduction in search result quality is likely to drive searchers to competitors. Moreover, in a market characterised by cutting-edge technology and a high degree of innovation, dominance is difficult to sustain; major innovation by a new entrant or an existing competitor can result in the loss of dominant position. Competitors that do not innovate and constantly improve consumers’ product experience get left behind. The history of the internet search market confirms this risk: Yahoo! was the market leader until Google surpassed it by providing consumers with more useful search results.8 Therefore, search engines necessarily have to provide best possible quality to their customers in order to maintain their market position. Reducing the quality of search results—or even merely failing to innovate as competitors continue to do so—would lead to searchers abandoning Google: because of the two-sided nature of the platform, eventually this leads to advertisers moving on as well.
The economics of online search engines mean that any strategy that tries to extract short-term benefits but harms the platform’s consumers in the long run would be irrational. It is therefore likely that Google’s innovation on built-in features was done to make search easier and more effective for its users. This, in turn, implies that there is no cause of concern from a competition law standpoint. In closing the case against Google, the US Federal Trade Commission (FTC) has noted that,
Product design is an important dimension of competition and condemning legitimate product improvements risks harming consumers. Reasonable minds may differ as to the best way to design a search results page and the best way to allocate space among organic links, paid advertisements, and other features. And reasonable search algorithms may differ as to how best to rank any given website. Challenging Google’s product design decisions in this case would require the Commission—or a court—to second-guess a firm’s product design decisions where plausible procompetitive justifications have been offered, and where those justifications are supported by ample evidence.9
If Google were not allowed to innovate on the built-in features, obviously, alternative search engines that directly compete with Google would benefit since they could manage to attract those Google consumers who have valued the built-in services. Therefore, any action against Google based on this allegation is akin to protecting competitors and not necessarily the competitive process.
Based on the discussion on the economics behind two-sided markets, it is clear that further analyses are required before arriving at a definitive conclusion in the Google case. A step-by-step analysis, which focuses on understanding appropriate market definition, and the subsequent competitive effects, is needed. Since the economics of two-sided markets differs significantly from the traditional markets, one needs to be cautious because some of the practices that may seem anticompetitive in traditional markets are indeed the standard market practices and equilibrium responses in these markets (for example, pricing below marginal cost in one side of the platform). Proper analysis of two-sided markets requires identifying the consumers on each side of the market, evaluating how a platform competes for each group of consumers, and determining whether a practice harms competitors in light of the fact that platforms create economic value through the connection they enable between groups of consumers. A lack of proper understanding of the two-sided nature of the markets can lead to erroneous conclusions that could harm consumer welfare.
One of the essential features of the online sector is fast-paced innovation. In particular, search providers like Google constantly keep innovating and changing all the time in order to adapt to changing circumstances. It is quite possible that several market dynamics—competitive practices, practices adopted by various players, etc—would have changed considerably between the time the complaint was originally filed and now. It is possible that some of the allegations, and hence the regulatory intervention by the antitrust authorities, may no longer be relevant. In order to check this, the CCI needs to include another layer to its analysis to review the markets and ensure that the competition landscape and the practices adopted have not changed significantly since the filing of the complaint.10
A pertinent question to ask in the context of competition law is, whether consumer harm has been perpetrated by the alleged conduct. In an industry characterised by low switching costs and fast-paced innovation, it becomes difficult for any firm (search engine) to undertake any activity that harms consumers. Moreover, it is also difficult to argue how built-in services—which provide an additional convenience to the consumers—can perpetrate consumer harm. Perhaps, because of this reason, in recent times, competition authorities across various jurisdictions have decided to close investigations against Google (for similar allegations to those that were filed in India) without action. Four specific examples highlight this. First, in its closing order, the US FTC observed that changes to Google’s search engine were legitimate product improvements made to improve Google’s search engine service and thereby strengthen the connection between the searchers and advertisers that use Google’s search engine advertising platform. Second, as pointed out earlier, the United Kingdom (UK) court has also arrived at a similar decision in response to the allegations by Streetmap, a firm that specialises in providing online maps in the UK. Streetmap has alleged that the presence of Google Maps in Google’s search results has led to Streetmap’s losing business, and hence, Google has abused its dominant position. While ruling in favour of Google, Justice Roth has concluded that, “I find that the introduction of Google in the UK … was not reasonably likely appreciably to affect competition in the market for online maps.”11 Finally, a German court in 2013 and the Taiwanese Fair Trade Commission (TFTC) in 2015 have also arrived at similar conclusions in their respective Google cases. In fact, the TFTC has also decided to close its investigation against Google (Wang and Wang 2016).
It is, therefore, relevant to ask whether there is anything about CCI’s investigation in India that would require a different result. The arguments used by the authorities that have closed their investigations are related to the nature of the two-sided markets in general and not necessarily restricted to any particular geography. Therefore, an important question to ask here is, are there any aspects of these two-sided markets that are different in the Indian setting when compared to these other countries? In other words, are there any systematic differences between the Indian search engine market and the markets in these countries that make these arguments less forceful in India? None of the arguments made above are contingent on any specific geography. Instead, they all focus on how a generic market operates. Therefore, it is quite difficult to imagine there being exclusive features of the Indian market that make these theories invalid. The only question that remains then is legal in nature: are there any specific features of the Indian competition law that differ from these jurisdictions significantly?
In February 2018, the Competition Commission of India (CCI) fined Google ₹135 crore for indulging in search bias. Google has since then appealed to the National Company Law Appellate Tribunal (NCLAT).
Notes 1 Platforms can be multisided, having more than two sets of constituents. Similar principles may apply where platforms have more than two sides. However, as explained later in the paper, search services are not multi-sided: there are only two sets of constituents (users and advertisers) and not three (the sites that may appear in search results).
2 For a detailed discussion see Anderson and Gabszewicz (2006).
3 In a traditional market, pricing significantly below or above marginal cost of production raises antitrust concerns. It could be looked upon as exploiting monopoly power or indulging in predatory pricing. In the platform markets, however, price below marginal cost (almost free) for one side is a perfectly valid business practice.
4 “Auction,” Setup and Basics, AdWords Help, viewed on 28 September 2017, https://support.google.com/adwords/answer/142918?hl=en.
5 In addition to simply attracting more users, search engines can also make their platform more attractive to advertisers by appealing to users that are especially likely to click on a given ad. For example, a specialized search engine for travel information may have relatively low traffic, but the chances of a given user clicking an ad related to travel is higher. See Varian (2006).
6 In fact, some market reports have suggested that display advertising with websites like Facebook and Twitter is likely to overtake internet search advertising by 2016. See, Lunden (2014).
7 Similar allegations have been raised in other jurisdictions as well, with the bone of contention being whether or not built-in services are standalone products or just product improvements aimed at improving the search results. For example, recently, Streetmap—a mapping services company in the United Kingdom—has alleged that Google’s practice of bundling Google Maps with Google Search has resulted in significant losses for itself. The Honourable Judge Roth of the High Court in Chancery Division has observed that such practice did not constitute abuse of dominance as alleged by Streetmap. See, Streetmap EU vs Google Inc.  EWHC 253 (Ch).
8 The Social Media space provides several examples where market leadership position was ceded to a superior player. For example, Facebook overtook MySpace in terms of popularity. In India too, Google’s Orkut was one of the leading social networking website before Facebook overtook it in terms of popularity. Google had to eventually close Orkut. For more examples, see Pingali (2016).
9 See “Statement of the Federal Trade Commission Regarding Google’s Search Practices in the Matter of Google Inc. FTC File Number 111-0163,” p 3, viewed on 26 September 2016, https://www.ftc.gov/system/files/documents/public_statements/295971/130103googlesearchstmtofcomm.pdf.
10 Dropping of antitrust cases because of significant changes in market landscape is not without precedent. In 1981, the United States government dropped its monopolisation case that it filed against IBM in 1969.
11 Case Number: HC-2013-00090, Streetmap vs Google Inc (Paragraph 177).
Anderson, Simon and Jean J Gabszewicz (2006): “The Media and Advertising: A Tale of Two-Sided Markets,” Handbook of Economics of Arts and Culture, Ginsburgh and Thorsby (eds), pp 567–614.
Edlin, Aaron and Robert G Harris (2013): “The Role of Switching Costs in Antitrust Analysis: An Analysis of Microsoft and Google,” Yale Journal of Law and Technology, Vol 15, No 2,
Evans, David (2010): Platform Economics: Essays on Multi-Sided Business, Competition Policy International.
Evans, David and Richard Schmalensee (2014): “The Antitrust Analysis of Multi-Sided Platform Businesses,” International Handbook of Antitrust Economics, Roger D Blair and
D Daniel Sokol (eds), Vol 2, Oxford: Oxford University Press.
Financial Times (2014): “India to Have 2nd Largest User-Base after China by 2016,” Financial Times, 23 November, viewed on 22 May 2018, https://www.financialexpress.com/industry/technology/india-to-have-2nd-largest-online-user-base-after-china-by-2016/11763/.
Ganz, Kian (2015): “Why Did CCI Write Google a Bad Report Card?” Livemint, 16 September, viewed on 22 September 2016, http://www.livemint.com/Companies/5D4c8f9kKB41IyL99Rfm4H/Why-did-CCI-write-Google-a-bad-report-card.html.
Gupta, Deepali (2015): “CCI Charges Google with Rigging Search Results; Flipkart, Facebook Corroborate Complaints,” Economic Times, 31 August, viewed on 22 September 2016, http://articles.economictimes.indiatimes.com/2015-08-31/news/66070945_1_google-finance-chairman-ashok-chawla-search.
Hindu (2013): “India Is Now World’s Third Largest Internet User after US, China,” Hindu, 24 August, viewed on 22 September 2016, http://www.thehindu.com/sci-tech/technology/internet/india-is-now-worlds-third-largest-internet-user-after-us-china/article5053115.ece.
Lunden, Ingrid (2014): “Internet Ad Spend to Reach $121B in 2014, 23% of $537B Total Ad Spend, Ad Tech Boosts Display,” TechCrunch, 7 April, viewed on 22 September 2016, https://techcrunch.com/2014/04/07/internet-ad-spend-to-reach-121b-in-2014-23-of-537b-total-ad-spend-ad-tech-gives-display-a-boost-over-search/.
Ohlhausen, M (2013): “Statement of Maureen Ohlhausen, in the Matter of Google Inc (FTC File No 110-163),” 3 January, viewed on 23 September 2016, https://www.ftc.gov/sites/default/files/documents/public_statements/statement-commissioner-maureen-ohlhausen/130103googlesearchohlhausenstmt.pdf.
Pingali, Viswanath (2016): “Are User Shares Indicative of Market Power?” LiveMint, 13 April, viewed on 27 September 2016, http://www.livemint.com/Opinion/nd41DmMEW3yIie8Ov3PxWN/Are-user-shares-indicative-of-market-power.html.
Ratliff, James and Daniel R Rubinfeld (2010): “Online Advertising: Defining Relevant Markets,” Journal of Competition Law and Economics, Vol 6, No 3, pp 653–86.
Rochet, Jean-Charles and Jean Tirole (2003): “Platform Competition in Two-sided Markets,” Journal of European Economic Association, Vol 1, No 4, pp 990–1029.
Rysman Mark (2009): “The Economics of Two-sided Markets,” Journal of Economic Perspectives, Vol 23, No 3, pp 125–43.
Salinger, Michael and Robert J Levinson (2015): “Economics of FTC’s Google Investigation,” Review of Industrial Organization, Vol 46, No 1, pp 25–57.
Varian, Hal (2006): “The Economics of Internet Search,” Rivista di Politica Economica, November–December, pp 177–91.
Wang, Su-Wan and Elizabeth Xiao-Ru Wang (2016): “Focus on Innovation: A Review of the Taiwan Fair Trade Commission’s Investigation on Google Maps,” Competition Policy International, July.