«Information, Privacy, and the Internet: An Economic Perspective Susan Athey Stanford Graduate School of Business1 Contents Introduction 5 1. ...»
and the Internet
An Economic Perspective
Susan Athey (Stanford, Microso )
Information, Privacy, and the Internet:
An Economic Perspective
Stanford Graduate School of Business1
1. Innovation, Value Creation and Data-Driven Advertising 11
2. Competition and Welfare in Search and Online Advertising 19
3. Information Gatekeepers, Data and Competition on the Internet 31
4. The Impact of User Multi-Homing on Publishers 43
5. An Economic View on Privacy 46 Conclusions 61 Introduction The arrival of the “era of big data” has been heralded as transformative for industry, economic growth, and efficiency (Mayer-Schönberger and Cukier, 2014). Every day brings new headlines about technological advances that have the potential to greatly improve our lives. Computers learn to understand our speech, e-commerce and mobile application shopping platforms present us with prescient recommendations, news about our friends and their babies is prioritized and ranked by social media sites, our travel reservations are organized into itineraries, and we listen to music curated by an algorithm. We can find inspirational photos that provide design ideas for remodeling our homes that previously would have been available only through professionals, and we can shop for homes online using maps, dozens of photos, and detailed information about homes, all neatly organized for us.
We have come to expect these services for free or at very low cost, and we are not at all surprised to see new websites, fresh news coverage, innovative mobile applications, and improvements in speech recognition every day.
We also notice advertisements that are increasingly tailored to our interests, with slideshows of products appearing in display ad windows. After browsing for autos, we begin to see more ads for cars. And advertisements begin to appear in places that were previously ad-free. Services that we have become attached to begin to replace free and unbiased content with paid content, and to promote their own complementary services. Others are acquired by the largest tech firms, or disappear in the face of an acquisition of a competitor by a large firm.
Although we always knew that our data was out there being used, it has become more salient as we see companies use it in new innovative ways. Yet, many of us are happy to hand over personal data in exchange for valuable services. For example, we might use an application that scans our email looking for travel plans and creates travel itineraries, saving valuable time. We might be delighted when the “personal assistant” function on our mobile phone reminds us of appointments. On the otherhand, news of widespread government surveillance has made many nervous, as they had not expected that this surveillance was occurring.
Surveillance has captured the imagination of the public, causing us to focus in a new way on the kinds of data companies have about us and how the data may be used in ways we did not originally intend.
As Microsoft’s Craig Mundie commented, “People are now being observed in increasingly intimate ways by all the technology in their lives. Too much data is being collected in too many ways” (Burt, 2013).
In the face of all of these trends, it is difficult for policy-makers to know how to respond. Antitrust and privacy regulators both face a classic tradeoff between allowing technological innovation to proceed disciplined only by the market and consumer choice, or intervening and risking doing more harm than good. Despite the fact that policy-makers profess an acute awareness of these tradeoffs, much of the public discourse around these issues misses fundamental facts as well as economic principles that provide some guidance as to when market solutions are likely to well and poorly, as well as when regulatory solutions might do well and poorly.
For example, conventional wisdom says that when markets are more competitive, social welfare generally increases, as competition induces firms to increase quality, lower prices, or both.
However, this wisdom only applies when the firms’ primary strategic alternatives are to increase quality or lower prices.
Technology platforms, as complex entities, often have a wider array of strategic choices. Many of the most damaging antitrust accusations against technology firms have occurred in a context where the firms in question were dominant but perceived the new arrival competitive threats, indeed threats that could dislodge their core business. For example, the Microsoft antitrust case in the 1990s revolved around Microsoft’s response to what it perceived as the threat of the internet browser—Microsoft was accused of promoting its own browser at the expense of others, using its dominant position in operating systems to gain leverage. In the 1980s, when American Airlines owned the Sabre reservation system for ranking flights, it was accused of responding to price competition and entry by low-cost carriers by ranking its own connecting flights above competitors’ nonstop flights in the flight reservation rankings, hoping to drive out new entrants by depriving them of customers. The current EU investigation of Google has focused on its manipulation of the search results page to promote its own specialized search sites at the expense of specialized competitors who threatened to compete away users in the most profitable segments of internet search.1 Despite these See Edelman (2014) for more examples of potential competition policy violations by Google.
prominent examples, it is tempting for regulators to simply argue that the technology sector is competitive and ignore the fact that not all competitive responses are welfare-enhancing.
This suggests that relying solely on market forces to achieve outcomes that benefit consumers may be short-sighted. At the same time, the challenges of creating regulation and policy that does more good than harm remain. Particularly difficult challenges emerge when creating regulatory policy for an industry with superdominant firms, because consumers may have to give up a fair bit in terms of their satisfaction with a product in order to choose a niche competitor with more attractive privacy policies.
In this paper, I will provide a selective review of some of the economic issues surrounding innovation on the internet, the role of data, and privacy. A deeper understanding of these issues is crucial to creating intelligent policy that achieves an appropriate balance between costs and benefits in a complex and highly innovative industry.
Some key insights highlighted in the paper include:
Consumer data creates enormous value for both consumers and advertisers, as measured by user engagement Value must be accrue to content creators rather than ad platforms and intermediaries, otherwise innovation will be hampered, decreasing the quantity and quality of future of internet content such as investigative journalism, low-cost or free productivity software, and entertainment Competition in advertising platforms is crucial to ensure that the value created with consumer data accrues to internet publishers and advertisers (who are often themselves internet content creators), rather than advertising platforms and intermediaries.
Advertising also plays another crucial role for new ventures. It is one of the only ways for businesses to be discovered. For example, a new application for a mobile device may have a difficult time rising to the top of the rankings without obtaining some initial users (and their positive rankings, as well as feedback) first. As internet search results pages for commercial queries leave less and less space for “algorithmically determined” links, new websites may rely heavily on advertising to be discovered. Only after some initial users are gained can the websites hope to rise to the top of the rankings to get new users organically. Targeted advertising can allow application developers to find an audience of users that may be interested in a niche application; without targeting, advertising may be prohibitively expensive for the new business, as much of the advertising expenditure may be wasted.
Thus, even though a naïve view of advertising would consider it a negative for consumers, it is hard to imagine life today without all of the consumer benefits that are directly enabled by advertising.
This has never been more salient than in the internet era. For this reason, the efficiency and effectiveness of advertising is of firstorder importance to determining consumer welfare from digital products and information.
In addition, advertising platforms and exchanges play a crucial role.
These institutions match publisher inventory to advertisers, and keep a share of the advertising revenue generated as profit for the ad platform. It is crucial for publishers, app developers, and content creators that ad platforms and intermediaries do not extract too high a share of the surplus created by advertising. That is, ad prices should not be artificially high, or new websites will not be able to afford the acquisition of new users; and the share of ad revenue given to publishers should not be too low, otherwise the advertising revenue will not be sufficient to incentives publishers to enter and create content. Of course, one way to discipline advertising platforms is to have competition in ad platforms.
Overall, then, the vibrant web and publishing ecosystem consumers enjoy today is crucially dependent on efficient and effective advertising, as well as competitive behavior by ad platforms.
How is Data Used to Create Value in Online Advertising?
Online advertising has created a variety of new and innovative forms of advertising, some of which is highly specialized to the context of the user. By far the most successful to date has been search advertising. In its pure form, a user enters a query and the search engine returns a set of “algorithmically ranked” links as well as a set of links to sponsored content, sold using a real-time auction. One reason that search advertising is so successful is that the advertisements are very similar to the non-paid content that the user is receiving. The user entered the search query in order to receive links to websites that had relevant information or sold relevant products. The advertisements also fill the same need, but are selected according to different criteria, where relevance is still a crucial component of how ads are chosen. Paid classified ads are another historical example where the advertisements are the content: classifieds make users aware of products or services that they might want to buy, and users go to the classified section to find that information.
Thus, search advertising is not purely wasteful nor entirely disliked by consumers; instead, it creates value. A profit-maximizing search engine will typically display more ads than would be ideal from a pure user experience perspective, but will rather trade off user experience and advertising revenue in choosing how many ads to display. So while the value creation is on average lower from ads than algorithmic results given the choices made by search engines, there is still a lot of value creation from the ads (as measured through, for example, click-through rates).
At the other extreme, classic “display” advertising was less tailored to the user’s context or intent. As a result, prices for display advertising were a small fraction of those in search. This advertising created less value to consumers and advertisers, and less revenue to publishers.