Vili Lehdonvirta, Oxford Internet Institute, University of Oxford
Cars were smashed and tires burned in France last year in protests against the ride hailing app Uber. Soon after an angry crowd attacked Uber cars with bars and stones outside Mexico City airport. After that, more than a thousand taxi drivers blocked streets in Rio de Janeiro. Despite protests such as these, Uber is very popular among its users and its usage keeps growing. Uber says it has over a million users in France and is now available in 71 countries and 429 cities. Uber is of course only the most visible and controversial of an entire generation of apps and online platforms that mediate between buyers and sellers of goods, services, and labour. Others include Airbnb, Taskrabbit, Freelancer, Upwork, 99designs and Shiply. They also share many design features with popular older platforms such as Ebay, Alibaba, and Amazon Mechanical Turk.
Unsurprisingly, policy makers at the highest levels are starting to get interested in platforms. A recent EU Commission flagship strategy paper notes that “online platforms are playing an ever more central role in social and economic life” (European Commission 2015). American policy makers have held several workshops and hearings on the topic. Undoubtedly there is a lot of unsubstantiated hype in the air about platforms. Many projections of their growth will turn out to have been overoptimistic. But even if we set aside future projections and look only at what platforms like Uber and Airbnb have already achieved, it is clear that platforms must be a topic of interest for policy makers and the social scientists who inform them.
There are two major theoretical literatures on platforms today. One is found at the intersection of technology management and industrial economics. In this literature, which we shall refer to as the economic literature, platforms are understood as products, services, firms, and institutions that mediate between two or more distinct types of agents (Rochet and Tirole 2003, Boudreau & Hagiu 2009). The quintessential example in this literature is a computer operating system. By defining a standard interface between software developers and users, the operating system coordinates their investments (Gawer 2009). Software developers can be assured that there is an audience for the products; users can be assured that there is software for their computer. A key concern in this literature is how a platform can solve its inherent chicken-and-egg problem and grow by cross-subsidizing the different sides (Parker & Van Alstyne 2005). A related concern is how platforms compete against each other in the presence of network externalities that tend to give rise to winner-takes-all dynamics (Noe & Parker 2005, Eisenmann et al. 2006).
The distribution of complementary software products has since moved from brick-and-mortar stores to app stores and other online distribution platforms. Online platforms have also started to mediate exchanges in other goods and services. We can refer to these new platforms as ‘online exchange platforms’ to distinguish them from the more general concept of platforms in this literature. Besides coordinating agents in the sense of ensuring compatibility and availability, online exchange platforms also coordinate agents in the very specific sense of matching individual buyers with sellers, via features such as search and recommendations. The economic literature on platforms has started to adapt to account for the distinct characteristics of these new platforms. Among other things, it has started to take influences from market design, a more recently emerged field of economics (Levin 2011, Vulkan et al. 2013). As a result, key concerns in the literature have expanded to the efficient matching of agents and the regulating of agents to eliminate undesirable behaviours (Parker et al. 2016).
The economic platform literature is very useful in addressing certain policy questions. For instance, it shows that platforms can provide services at a loss to one side of the market whilst still making an overall profit, challenging conventional wisdom on predatory pricing and thus potentially improving competition policy (Parker & Van Alstyne 2005, Rochet & Tirole 2003). It has less to say about urgent policy questions pertaining to platforms in the context of the broader society. Parker and colleagues (2016) provide a very good overview of such issues, ranging from taxation and labour rights to state sovereignty. But they do not offer any particular framework for analysing these issues and deriving policy conclusions. This task is beyond the conventional scope of the economic platform literature, requiring broader theory.
Another major literature on platforms is situated in media and communication studies. It is more recently established and as yet smaller in volume than the economic literature. But where the economic literature zooms in on dynamics inside and between platforms, the communications literature considers platforms in the context of the broader society and international system, usually from a critical political economy perspective (Gillespie 2010, Dijck 2013, Jin 2015). A key premise is that an increasing share of the social, economic, and political interactions that constitute society are mediated by large online platforms (Dijck 2013). Platforms are thus in a powerful position to regulate parts of society, and in doing so take on functions comparable to the state (Jin 2015). They can be a countervailing force to the state in areas such as privacy, surveillance, and censorship, or a complicit partner that enhances state power and extends it over new territory. Platforms can also benefit from the “free labour” of their users (Jin & Feenberg 2015) or the commodification of everyday life into marketable forms (Dijck 2013).
The communication literature is vague when it comes to what exactly constitutes a platform. Indeed, one of its key claims is that the term ‘platform’ is deliberately vague in meaning so as to allow platform companies to present themselves as both neutral conduits and private curators as political expediency requires (Gillespie 2010). In practice, the literature has focused on social media platforms and other digital information and communication platforms, such as Google Search. It has paid less attention to what we have termed as online exchange platforms. The literature has been very good at problematizing issues related to platforms and creating a counter-narrative to the techno-utopian narratives of technology companies and tech media. Where it falls short is in theorizing what value platforms can provide and in generating concrete answers to urgent policy questions around online exchange platforms. It fosters critical inquiry, but provides few tools for developing specific policy responses and assessing potential impacts.
Addressing the policy questions around online exchange platforms (henceforth just ’platforms’) would seem to require a theory framework that combines some of the broad sweep of the communication literature with some of the reductionism of the economic literature. It would be comprehensive enough to cover the issues of interest yet sufficiently stylized to yield specific conclusions. Fortunately there is a major theoretical framework that fits this description: New Institutional Economics, including especially its elaborations in economic history and political science.
Platforms’ rapid expansion means that public policy makers at all levels are confronted with some basic, urgent questions. Should a platform be permitted to operate in this jurisdiction or not? If yes, how should it be regulated? Must we accept that the greatest positive network externalities are obtained when global platform companies take over the entire market, or should we try to promote local entrepreneurs or even publicly operated platforms? The aim of this paper is not to provide any conclusive answers to these questions, but to sketch a theoretical framework for thinking about platforms in the context of the wider economy. This will allow answers to be generated with further research into specific contexts and cases.