Polity Press, Cambridge, UK and Malden, MA, 2016. 120pp., £9.99 pb
Reviewed by Bruce Robinson
Bruce Robinson is a retired lecturer in Information Technology and a socialist activist. He is currently researching the relationship between Marxism and artificial intelligence and the radical history of Clerkenwell.
Uber or Airbnb are regularly making the headlines as a result of their business practices. At the same time, Facebook, Amazon and Google are now presented, whether positively or negatively, as paradigms of a new form of capitalism. These firms all operate on the basis of the platform, the digital infrastructure that enables profits to be made from providing the connectivity that enables business transactions and social interactions to take place through the Internet. A business model has developed based around the new capacities of platforms to use data gathered by means of the interactions that take place through them.
This phenomenon – platform capitalism – is the subject of this short book by Nick Srnicek, co-author of #Accelerate: Manifesto for an Accelerationist Politics and Inventing The Future: Postcapitalism and a World without Work. It aims to give ‘an economic history of capitalism and digital technology while recognising the diversity of economic forms and competitive tensions inherent in the contemporary economy’. (3) It takes as its framework a stance that is anti-capitalist without being explicitly Marxist. The book has three chapters which set out the past, present and future of platform capitalism, focussing on the political economy rather than the technology or the social consequences, which means that there is deliberately little discussion of its implications for labour. Yet the worldwide worker resistance to the encroachment of Uber and proposals for the legal regulation of platform-based firms point to this as a lack of an important dimension that may have a significant influence on the future of platform capitalism.
Srnicek begins with a short history of capitalism since 1945, moving from the post-war settlement to its crisis in the 70s. He argues that manufacturing faced low profitability and excess capacity, which forced capital to look elsewhere for profitable investments. This initially resulted in the dot com boom and bust of the late 90s, which resulted in a privatised Internet infrastructure together with a period of low interest rates. An intensification of this monetary policy after the crisis of 2008 reduced the appeal of financial assets, encouraged riskier investments, and ‘was a key enabling condition for parts of today’s digital economy to rise.’(27) This was boosted by massive corporate cash hoarding, particularly expanding the coffers of tech companies.
Where does the technology of the platform fit into this account? ‘With a long decline in manufacturing profitability, capitalism has turned to data as one way to maintain economic growth and vitality in the face of a sluggish production sector… The platform has emerged as a new business model, capable of extracting and controlling immense amounts of data.’ (6) The platform becomes the means to connect users (e.g. Facebook friends, Airbnb property owners and their renters) and, as their activities take place through the platform, it can record and put to use data about them. (44)
Data is then the raw material on which this advanced sector of capitalism is centred (39), whether it serves as a commodity in itself or as a means to gain competitive advantage. While data has become increasingly important in the functioning of capitalist enterprises, I am sceptical of the book’s argument that investment in tech firms has been driven by data rather than the more general search for profitable outlets for financial capital. The ‘turn to data’ is perhaps more fortuitous than is implied. Speculative capital has flowed into many tech firms – profitable ones such as Google included – before the mechanism for profitable operation was decided. Investment preceded monetisation. Successful monetisation can still be problematic as is shown by Twitter’s continued lack of profitability despite its popularity. Srnicek acknowledges the problems of sustaining advertising revenues – the vast majority of revenue for Facebook and Google – and sees it pushing firms towards diversification. However he still centres his analysis on data as the motive force of platform capitalism.
Srnicek employs a useful typology of platforms which enables us to see not merely the heterogeneous applications of the platform model and their differing uses of data but also their differences in terms of the processes of value production and capital accumulation. He distinguishes five types of platform, each of which is distinct from but draws on non-platform based predecessors.
The first and best known platform model is the advertising platform where the provision of services is used to collect data that enables the users to be the targets of advertising tailored to them, thus cutting costs and increasing the effectiveness for the advertiser. This model has become the standard for platforms providing mass free Internet services. It is based on the accumulation of user data.
Cloud platforms essentially involve the renting out of IT resources and services by firms that have developed platforms for their own use with the result that user firms can effectively outsource their IT department. Amazon developed a vast platform to support its own business and through it makes available to others cloud computing including computing power, storage, software development tools and packaged software. It has been estimated to be worth $70 billion and competitors such as IBM, Google and Microsoft are entering this market. It enables user firms to cut their capital costs by employing these services on an ‘on demand’ basis. (This is not a new development. Renting computer time and storage and packaged software have been around since the age of the mainframe.)
Industrial platforms represent an attempt to apply the platform model to industrial processes using the ‘Internet of things’ so that chips and sensors embedded in products can enable coordination via a platform that gathers the data and controls the production process. This is being extended from manufacture to include design and ‘mass customisation’ of products in response to demand. Like previous methods such as Enterprise Resource Planning software, industrial platforms aim to integrate the circuit of capital increasing the speed of capital turnover, cutting labour and capital costs and making possible rapid responses to shifts in demand.
Product platforms are essentially platforms that enable the payment of a rent for the use of a product that remains owned by the platform owner, the producer or another intermediary who has a claim on the product. Spotify collects the rent on its recordings as an intermediary passing some of it back to record companies who may in turn pass on some of the revenue to the maker of the music. This model has expanded to include means of obtaining high cost consumer purchases such as cars on a rental basis through to the renting of capital goods such as railway rolling stock or jet engines where Rolls Royce pioneered a model based on charging fees for the time an aircraft engine is used managed by a platform that gathers data on engines’ use.
Lean platforms, exemplified by Uber and Airbnb, are distinguished from product platforms in that they ‘attempt to outsource nearly every possible cost’ (71) so that the intermediary role of the platform itself becomes the use value that attracts users and the means to lay claim to a share of the revenues accruing to the taxi driver or home owner. All that remains in-house it is ‘a bare extractive minimum – control over the platform that enables a monopoly rent to be gained’ (76). Srnicek sees this as ‘simply extending earlier trends [of outsourcing] into new areas’ such as small intellectual tasks.
Lean platforms are central to the ‘gig economy’. Srnicek points to the implications for outsourced labour in terms of casualisation and a race to the bottom for non-place-dependent tasks ‘expanding the labour supply to a near global level.’ (82) He quotes figures of around 1% of the US labour force working through lean platforms with about 1.3 million UK workers working through them once a week.
These very different applications point to the centrality of data collection in the platform model. If data is a raw material, requiring further analysis and processing, then a raw material for what? What’s immediately clear from the list of functions this data supports (40-1) is that, though data can itself be a commodity, it often plays a secondary role in enabling or improving other economic functions. This then raises the question of whether a data-driven boom is sustainable without expansion in other sectors of the economy. Srnicek answers in the negative (56), rightly pointing to its inability to overcome the fundamental issues of capital accumulation as he defines them and suggesting that these firms may be ‘parasitical on other value-producing industries’.
This requires further analysis in terms of the specific roles of different kinds of data in value production and circulation as some are productive and others unproductive in Marx’s sense and some produce profit while others produce rent. If, as Srnicek concludes, ‘platform capitalism has inbuilt tendencies to move towards extracting rents by providing services’ (126) , do platforms represent a shift towards a rentier capitalism, the ‘becoming rent of profit’? The book does not comment on this debate but I would instead argue that the types of platform Srnicek mentions in this context (‘cloud platforms, infrastructural platforms and product platforms’) represent a simple contracting-out of the functions of capital so as to mitigate the costs or raise the speed of turnover of capital and remain within the straightforward circuit of capital of individual processes of production. This form of contracting out is distinct from lean platforms that simply slice a part of the value from transactions that would otherwise take place by different means.
Srnicek dips his toe into the ‘digital labour debate’ rightly rejecting the position of Christian Fuchs and autonomist writers that the ‘free labour’ of the users of Internet services is productive of value. Instead ‘data is appropriated as a raw material’ (53-6) where appropriation occurs through the imposition of legal and technical terms and conditions for the use of Internet services. This is part of a drive to expand control of all Internet-mediated activities flowing from the specific competitive pressures facing platform-based companies.
The book’s final chapter entitled ‘Great Platform Wars’ looks at the future prospects of platform capitalism based on a number of tendencies which are built into the platform as a means of production and communication. Above all, platform enterprises share a tendency towards monopoly as a result of network effects – broadly, ‘the more numerous the users who use a platform, the more valuable it becomes for everyone else’ (45). This benefits both users who gain access to more connections – whether Facebook friends, taxi drivers or rooms – through the platform and the platform owners who can gather more data. ‘Users beget more users, which leads to platforms having a natural tendency towards monopolisation’. (45) Thus Facebook becomes the place for social contacts, Google for search, Airbnb for room booking. The costs of growth are low. Srnicek also points to a ‘growth before profit’ strategy with a period of low or no returns as firms try to become dominant. (119) However some applications may never become profitable as the result of obstacles to making free services pay – which points to the glaring conflict between private ownership and social use value in this sphere.
Srnicek discusses a number of strategies which advertising platforms use to overcome competition and a decline in the growth of advertising revenues. The drive for data means the cross-subsidisation of free services, subscriptions or the introduction of payments alongside free services for a full or extended service, or the removal of advertising. The big monopolies expand in a way as to make increasing areas of interactions pass through their gateways, one way being through its own closed apps. For example, increasing numbers of people get their news through their Facebook connection and many other services now offer a seamless entry point via Facebook or Twitter. At the same time, the firms converge on similar areas in order to be able to keep users within their boundaries – each must offer cloud storage, a photo app and so on.
One of the virtues of this book is that it has a proportionate view of platform capitalism, neither seeing it as ushering in a new phase of capitalism (7-8) nor as portending its final crisis. Unlike his other writings, it is not based on an uncritical technological enthusiasm. His analysis is that the further development of platform capitalism faces serious challenges – through declining advertising revenues or the exhaustion of the potential of outsourcing – but that the already established monopolies can still expand into new areas and bolster their dominant positions.
The book devotes only a couple of short pages to discussion of how the platform model might be used outside the framework of encroaching capitalist monopolies. It rejects platform cooperatives as a solution on the traditional grounds that they involve competition and self-exploitation under capitalism, factors that are made worse by the network effects, access to resources and monopolistic control of existing platforms. Srnicek’s alternative in the book – to set up platforms as public utilities ‘owned and controlled by the people’ but supported by the state’s resources (128) – faces the same problem of network effects and competition as long as the existing platforms remain in private hands. Their collective ownership is thus a precondition for the democratic and socially useful development of the technology (as he acknowledges in a recent Guardian article).
The book stops short of thinking about the sort of platforms that could be useful in a ‘post-capitalist’ economy. Is it enough to remove advertising and surveillance from Google and Facebook or do the fundamental software structures that constrain their use need radical change? Can platforms become the basis for a democratically, planned economy by enabling the connection of consumers and producers without the mediation of the market?
‘Platform Capitalism’ provides a useful introduction to a dynamic sector of contemporary capitalism that has become directly influential in our everyday lives. It contains an extensive and up to date bibliography. However it is too short to meet its ambitious goal of being ‘an economic history of capitalism and digital technology’(3), although it covers many recent developments in depth and the material included in it is valuable. It also raises a number of questions which should be the subject of further research and debate in analysing the consequences of and alternatives to the continuing adoption of new technological forms by capital.
15 September 2017