Governance and Incentive Systems of Blockchain-Based Social Networks
* Article: The Common Factory: Governance and Incentive Systems of Blockchain-based Social Networks. By Felix Fritsch. Conference Paper, June 2019
CSC 2019 – 15th International Conference on Sociocybernetics Urbino, Italy – 25-29 June 2019 ; “Dark Ages 2.0”: Social Media And Their Impact Paper No. #43
Contextual Quote
"Blockchain is inconceivable without commodification – its very architecture rests on economic incentives coordinating independent actors, and its raison d’etre is the dis-intermediation of digital asset transfer. But their unique features not only allow blockchains to dis-intermediate and reorganize existing markets, but also to create new markets under previously unfavorable conditions. They thus constitute a crucial element in looming processes of formal subsumption, pushing the frontiers of capitalism ever further into the sphere of social relations previously unattainable to commodification. What’s more, through incentive systems rewarding specific interactions and punishing others, these technologies open the door for what Marx referred to as real subsumption – the systematic restructuration of commodified relations along the needs of capitalist profit maximization – of hitherto not or only formally subsumed social spheres."
- Felix Fritsch [1]
Abstract:
"Blockchain architecture provides robust institutional infrastructure for distributed networks that fundamentally changes the ways groups organize online. Its applications go beyond existing digital commerce to commodify and transform social relations through tokenization, incentivization and rationalization. Crypto Social Networks promising to reward users financially for valuable contributions constitute a paragon of such subsumption through blockchain. These platforms organize value distribution among users along affective metrics and introduce various mechanisms that facilitate autonomous group governance. Their emergence both acknowledges the critique issued by digital labour scholars of hitherto exploitative platform-user relations and challenges their uncritical appliance of time-based labour theory of value to the digital realm.
Beyond simple commodification, blockchains and the code they operate are discussed as emerging public infrastructure for the transaction of private assets and the decentralized organization of groups – as digital commons facilitating the total subsumption of sociality under abstract capital. While blockchain is inherently commodified and constitutes the capstone of a digital ‚social factory‘, the specific alignment of token design, incentive systems and other governance mechanisms can lead to vastly different outcomes in power and revenue distribution. The empirical section is devoted to highlighting these differences through analysis of three emerging platforms Minds, Steem and HyperSpace along their token, incentive and governance systems. While the findings underline the devil in the details especially of tokenomics, they also show the immatureness of existing platforms, which finds expression in different blind spots depending on the distinct means and aims of each project."
Discussion
Felix Fritsch:
"Referred to as foundation of the ‘Internet of value’, blockchain is inconceivable without commodification – its very architecture rests on economic incentives coordinating independent actors, and its raison d’etre is the dis-intermediation of digital asset transfer. But their unique features not only allow blockchains to dis-intermediate and reorganize existing markets, but also to create new markets under previously unfavorable conditions. They thus constitute a crucial element in looming processes of formal subsumption, pushing the frontiers of capitalism ever further into the sphere of social relations previously unattainable to commodification. What’s more, through incentive systems rewarding specific interactions and punishing others, these technologies open the door for what Marx referred to as real subsumption – the systematic restructuration of commodified relations along the needs of capitalist profit maximization – of hitherto not or only formally subsumed social spheres. Its closest advocates, by and large critical of monopolist power but not of commodification itself, rather point at the liberating aspects encryption and decentralized markets entail for the individual. Beyond these, they forecast potentially extensive impact of blockchain-based governance on the organization of social and economic relations. They hold that fully distributed, automated governance mechanisms embodied by (combinations of) smart contracts are not dependent on the continued goodwill of any benevolent dictator acting as central guarantor of rights and agreements, but become self-executing entities, insusceptible to corruption by external powers. Public blockchains further provide permission-less access to transactional data and open source code of both base protocol and decentralized applications, which fundamentally challenges ownership of these technologies. These qualities, so its proponents, position blockchain as foundational to an emerging structure of digital commons that come into being as new technological reality.
Blockchain-based applications are not confined to digital commerce in terms of peer-to-peer and multi-sided markets, but also cover collaborative social media platforms and with them their nonprofessional users (Glaser 2017). Already, a multitude of blockchain-based projects explicitly seek to challenge social media behemoths with decentralized platforms that leave users more control and reward them financially for their interactions. Such Crypto Social Networks (CSNs) make a particularly good case for the study of the effects of blockchain on society as they allow observation both of the subsumption of social relations under the reign of capital and of the constitution of selfsustained digital communities in a new digital commons." (https://www.researchgate.net/publication/347390602_The_Common_Factory_Governance_and_Incentive_Systems_of_Blockchain-based_Social_Networks)
From the Conclusion:
"This paper has highlighted the twist CSNs add to the prominent discussion over who produces the value social media platforms realize. Scrutinizing incentive systems along the forms and accounted units of activities recognized as valuable has shown limitations of the orthodox LTV in the realm of social media prosumption, and has affirmed the relevance of an affect-based reformulation of the LTV under conditions of immaterial production. The increasing importance of subjective valuation as opposed to ‘objective’ measures likewise shows in the speculative value of cryptographic tokens, which in the absence of market fundamentals is by and large based on affective brand value.
While acknowledging Marxist scholars’ point that prosumption constitutes digital labour that should be rewarded, the Crypto Social Networks under scrutiny all embrace commodification as their preferred means to overcome exploitation. Their promise to dissolve corporate platform ownership to the benefit of its users thus comes at the cost of total subsumption of sociality under the regime of an ever more omniferous social factory, as conceptualized and loathed by Italian Autonomists Mario Tronti (1966), Antonio Negri (1989), Nick Dyer-Witheford (1999) or Paolo Virno (2004). The resulting ‘common factory’, collectively controlled by its users, may overcome central modes of accumulation through exploitation, but it remains organized according to logics of productivity that pursue a totalitarian vision of Homo Economicus (Atzori 2015). At the same time as it cedes centralized control, the prerogative of capital becomes inscribed into a platforms’ very foundations, the blockchain infrastructure. And yet, the platforms built on top employ considerably different approaches to digitally engineering sociality (van Dijck 2012 and 2013; Atzori 2015; Atzori and Ulieru 2017) that result in specific constellations of power and revenue distribution. While not overcoming commodification as such, these differences are worth researching both for their impact on such constellations and for their wider transformative potential vis-a-vis centralized incumbents.
In light of these considerations, this paper may also be seen as the starting point of both a theoretical exercise to conceptualize ‘common factory’ as synthesis of social factory and digital commons and an empirical exploration of commonalities and differences of its implementation in dApps."
Does the Labor Theory of Value Function in Social Media ?
Felix Fritsch:
"The literature review provides an overview of the main lines of contention among critical scholarship over the persistence of Marx’ Labour Theory of Value (LTV) and the increasing relevance of rentseeking strategies in contemporary capitalism. In a prelude to empirical analysis, I assess the potential of different blockchain-based governance features for social media. Following this, three emerging CSNs are scrutinized in terms of their token economics, incentive systems and governance structures as instances of a phenomenon drastically changing relations among digital users and between them and the underlying platforms. In the discussion, the findings are summarized and interpreted in terms of their relevance for the theoretical approaches sketched out in the literature review. The conclusion highlights the twist CSNs add to contemporary debates on prosumption as they seemingly overcome exploitation at the price of total commodification. It closes with an introduction of a theoretical endeavor to conceptualize the conflation of social factory and digital commons in what I term ‘common factory’.
The Free Labour Debate
The emergence of interactive, or social, online media in the early 2000s came with celebratory accounts of the ‘electronic agora’ (Barbrook and Andy 2015) that activates citizens to participate in a horizontal, egalitarian digital public space free of censorship (critically reflected by Srnicek 2016; Fisher 2012; Fuchs and Sevignani 2013; Bengtsson 2013). Soon, however, materialist critique of the economic role of users started to form around Tiziana Terranova’s notion of ‘free labour’ (2000) to shed light on the exploitative relationship between users providing unpaid labour online and platforms accumulating surplus value. In an early contribution, Cote and Pybus analyze these new forms of immaterial labour as revolving around the biopolitical production of users’ subjectivity and highlight the role of affect “as the binding, dynamic force which both animates those subjectivities and provides coherence to the networked relations” (2007:90). They configure their work as a coeval supplantation of Dallas Smythe’s concept of ‘audience commodity’ that more explicitly takes an Autonomist stance and better accounts for the “qualitative shift in which culture, subjectivity, and capital come together in new networks of ICT” (Cote and Pybus 2007:98). This text is particularly crucial as it becomes a point of reference both for the emerging ‘free labour orthodoxy’, seeking to revive Smythe and combine the production of ‘audience commodity’ with ‘prosumer labour’ producing content and user data, and for scholars highlighting the role of affect as replacing labour-power as measure of value. The ‘free labour orthodoxy’ starts consolidating in the early 2010s with a comprehensive conceptualization of paid and unpaid digital labour as base for augmenting Marx’ theory of value to the digital realm. Scholars such as Fisher (2012, 2015), Fuchs (2010, 2011, 2012a, 2012b, 2014), Manzerolle (2010), Olsson (2013) and Ritzer and Jurgenson (2010) appropriate Alvin Toffler’s notion of ‘prosumption’ (1980) as combination of production and consumption, but use it in critical terms as “outsourcing work to users and consumers, who work without payment” (Fuchs 2011:297).
Constituting an essential refinement to the product of audience labour, prosumer labour draws on creativity and social relations of users that provide use value for other users and, in combination with “constant, total and algorithmic” (Fuchs 2015:54) surveillance, produce a rich data source for personalized advertising. In effect, prosumer labour, combined with audience labour, produces a targeted audience commodity, which is algorithmically auctioned to advertisers every time a user views or clicks on an advertisement. Considerable disagreement over the drivers and particularities of contemporary capitalism, the economic sphere advertising activities are located in, or the extent to which exploitation and coercion are applicable has been by and large resolved through continuous debate and refinement of conceptualization over the last years (e.g. Beverungen et al. 2015). The essence of the free labour orthodoxy is that Marx’ labour theory of value (LTV) also holds in the digital age, with the addendum that social media platforms appropriate significant amounts of surplus value from their users providing digital labour for free. Contentions based on Affect and Rents There are two main contentions with this convention. The first one builds on Antonio Negri and Michael Hardt’s (1999) considerations regarding the transformation of the LTV in postmodernity. Facing the increasingly ineffectual measure of labour-value in either use-value or money, they suggested to replace it with affect as ‘power to act’. Following their lead, Christina Morini and Andrea Fumagalli (2010) seek to adapt the LTV to new modes of accumulation under ‘bio-capitalism’. Going beyond notions of the knowledge-based economy, bio-capitalism incorporates “the entirety of human faculties, from relational-linguistic to affective-sensorial” (2010:235). Likewise, they identify multiple forms of ‘bio-labour’, named after their most significant, effective and difficult task: relational, cognitive, symbolic and affective labour. Analyzing the centrality of difference, subjectivity and contextualized knowledge, the two authors establish an inextricable link between cognitive capabilities and relational activities as “basic elements of the general intellect” (2010:238). Bio-capitalism outgrows Marx’ LTV as it dissolves the separations between work- and life-time, work- and life-space, and between production, reproduction, circulation and consumption – in other words, the walls of the factory: If all of life is labour, there is no outside to measure it against. Instead, Morini and Fumagalli propose a ‘life theory of value’ that acknowledges the multiplicity of new labour-forms as well as their differential, substantial exploitation in immaterial production. With regards to measurement of value, they hold that “the quota of added value (…) is dependent on supply, circulation and diffusion of immaterial commodities (by definition impossible to be numerically measured)” (2010:247f).
Adam Arvidsson and Eleanor Colleoni (2012), in a critical assessment and rebuttal of Fuchs’ (2010) argument on the relevance of Marx’ LTV for prosumption, likewise argue that value creation in informational capitalism is poorly related to time, especially in the newly commodified spheres of social life that are subject to prosumption. Instead, they hold that “value is ever more related to the ability to create and reaffirm affective bonds” (Arvidsson and Colleoni 2012:136). This also implicates to go beyond assessment of Facebook’s business model in terms of production and circulation of advertisement space and understand it first and foremost as financial venture that appropriates rent in the stock market. This is taken as instance of “an extended, society-wide process of finance-centered accumulation, where the link between reputational (or affective) value and access to financial rent becomes fundamental” (2012:136). Facebook is both subject to this general process and takes a special role as it serves the expression, objectivization and quantification of affect in the form of ‘likes’ as a new convention on the financial evaluation of stocks. Daniel Cockayne (2016) likewise locates the value of social media-generated big data in indicating user affinity for a platform itself. Especially early-stage platforms use these data vis-a-vis potential investors to demonstrate “the affective attachment prosumers show toward” (2016:8) them as base for future revenue generation. In more abstract terms, they present themselves “as an affective apparatus of capture” (2016:2) upon which future coercion into provision of free audience labour may build.
The second main contention disputes the qualification of prosumption as labour – both in a colloquial sense, as it belittles experiences of ‘real’ exploitation, and in terms of scientific Marxism (e.g. Arvidsson 2011). Its adherents uphold that users give their data deliberately and receive free services as a remuneration in kind and that raw data require skilled work to become valuable (Comor 2015; Srnicek 2016). The vast profits leading social media platforms generate in relation to their total expenditure for wages disqualify explanations that solely rely on direct exploitation of paid labour. Alternatively, some point to the market power of the largest advertising platforms as enabling the extraction of monopoly rent from customers who compete over market share through online marketing (Huws 2014). As the increased competitive pressure is at least partly forwarded to employees, some scholars also argue that their exploitation increases to allow for redirection of surplus value to advertising platforms (Rigi and Prey 2015). This argument tends to consider the advertisement sector part of the sphere of circulation and therefore per se not productive, but rather parasitically nourishing on the surplus produced in other spheres (Foley 2013; Huws 2014; Pasquinelli 2010). Other scholars have put more emphasis on big data flows gathered by platforms that they see as constituting a new, quasi-natural resource, which in Marxist terms generate rent when exploited. Platforms create the infrastructure to extract the data being created on them and refine them to build valuable data stock over time (Comor 2015). This stock is not sold itself but becomes the source of information that is sold as analysis to advertising customers (Srnicek 2016:31). Industry leaders have a monopoly insofar as their surveillance architecture is extremely detailed and, due to plug-ins on third-party homepages, almost omnipresent." (https://www.researchgate.net/publication/347390602_The_Common_Factory_Governance_and_Incentive_Systems_of_Blockchain-based_Social_Networks)
The Political Economy of Crypto Social Networks
Felix Fritsch:
"The explicit acknowledgment of hitherto exploitative platform-user relations makes a strong case for the critique raised by the free labour orthodoxy and against rents approaches explicitly rejecting notions of labour and exploitation. Given CSNs’ rejection of data stock accumulation through excessive surveillance practices, approaches focusing on data as a new resource likewise lose ground. What remains is the market power of dominant web 2.0 platforms, which for emerging web 3 platforms may constitute a restraining factor – if they base their business model on participation in that market. Rent may further be located in the interest (delegated) Proof-of-Stake consensus mechanisms offer; however, these are located on the protocol level of blockchains and, with exception of Steem, are indifferent to the kind of dApp running on them. Other than that, rents approaches do not seem to bear much relevance for the analysis of CSNs.
While CSNs acknowledge the free labour orthodoxy’s critique of exploitation, they refute its insistence on time as base of measurement. Incentive systems reward specific actions and reactions – they do not measure input in time but in instances. Their metrics relate to different forms of bio-labour – the cognitive, symbolic and affective labour enshrined in user-created content, the cognitive and emotional labour users apply when curating content, or the relational labour of sharing. The role of social media metrics in the evaluation of financial assets highlighted by Arvidsson and Colleoni (2012) shows in the prominent position leading CSN platform Steem has for circulating crypto market analysis. Their focus on increasingly finance-centered accumulation likewise reflects in the business model of the companies developing CSNs as autonomous entities, which predominantly relies on token value appreciation as source of profit. Cockayne’s depiction of platforms as “affective apparatus of capture” (2016:2) then opens the question how token value relates to a CSN’s ability to capture user affect towards advertisements. Given these cross connections, affect-centered approaches, particularly ones that acknowledge the exploitative dimension of prosumption, appear as promising advancement of the free labour orthodoxy in the age of blockchain.
The open source nature of public blockchains and thus of Smart Contracts deployed on them commands that copying and modifying these emerging institutional frameworks (Hayes 2019) comes at negligible cost. Competitive pressure towards their optimization (at least theoretically) forecloses any long-term rent-seeking strategies based on ownership or central control of this infrastructure. Institutions come to exist as dApps which cannot be enclosed and thus constitute new forms of digital commons. On the other hand, decentralized structures that possess internal capital – decentralized organizations, decentralized autonomous organizations and decentralized autonomous corporations (Buterin 2014) – potentially constitute “the enabling framework for the digital rule of capital: the total commodification of global society” (Garrod 2016:71). Atzori and Ulieru concede the former by referring to “the advent of blockchain-based digital platforms running on commons principles via self-enforcing contracts” (2017:2). However, they also acknowledge the latter as they warrant that under capitalist conditions, “digital platforms and the Internet at large cannot cease to be a medium of insatiable commercial appetites” (Ibd.:3) ingraining themselves in new organizational forms. They thus propose to scrutinize the specific architectural design of individual platforms in terms of decision rights, token issuance and value distribution through incentive systems to determine their relation towards contemporary capitalism. In an inquiry of the dApp ‘Swarm City’, Beck et al. (2018) likewise use decision rights, accountability and incentives as central analytical dimensions derived from IT governance literature and principal-agent theory.
Accordingly, this empirical investigation of CSNs as more or less owned, or at least to some degree controlled, by its users will focus on three governance mechanisms: Permission-less token issuance allowing the creation of post-corporate economies; Binding decision making tools giving network members power over rules and financial flows; and rewards schemes based on micro-transactions realigning incentives to support a common goal. The former not only allows project developers to raise initial funding and bind a core community, but also to construct such projects as fully decentralized economies not under their control nor generating permanent income flows for them, while still generating them profit from appreciating token prices. Decision rights depend on their distribution and can be differentiated into management rights to propose, execute, or implement decisions and control rights to ratify and monitor. Incentives encounter monetary and non-monetary rewards that motivate agents to act in predetermined and observable ways to obtain incentive alignment (Beck et al. 2018). In this regard, Atzori and Ulieru highlight the alignment of long-term interests through reputation games and inscription of “rights and privileges in social organizations through cooperative currency designs” (2017:7) as key to blockchain cooperatives that embody a ‘real’ sharing economy beyond platform capitalism. However, even in their ideal configuration, they warrant, the reliance of such platforms on incentive systems to shape human behaviour results in its rationalization and robotization and voids human interaction of spontaneity and much of its depth. What remains, even under conditions of post-capitalist collaborative exchange, is “the hyperrational, positivist mentality of the Homo Economicus – an “anthropological monster”” (Atzori and Ulieru 2017:12; see also Atzori 2015)." (https://www.researchgate.net/publication/347390602_The_Common_Factory_Governance_and_Incentive_Systems_of_Blockchain-based_Social_Networks)
Comparative Analysis of Three Crypto Social Networks
Felix Fritsch:
"The three platforms scrutinized above are distinct in the labour they reward, their openness to capital, the services their token stands for as well as size and origins of their rewards pool. While Minds focuses on the affective labour content producers provide, Steem and HyperSpace additionally reward curation as emotional labour of consumers. Affect towards content is formalized as discrete unit of account in all cases, however HyperSpace goes further in allowing users to determine the quantity of affect specific content solicits. Temporal measures only play minor roles: Minds rewards daily log in, HyperSpace restricts its daily token issuance to active users, and Steem likewise limits users’ voting capacity in temporal terms.
The interest rate Steem grants financial capital is rooted in its architecture as blockchain and thus cannot be compared with the other dApps. The central role it grants financial capital to leverage curation power and rewards, however, goes far beyond its competitors and makes Steem an attention plutocracy par excellence. HyperSpace’s design is most interesting in this regard, as its egalitarian distribution of voting rights seeks to actively counterbalance financial power. The lack of information regarding envisaged tokenomics and use value of tokens beyond currency, however, warrants against too optimistic appraisal of this model – its resolution could come with significant limitations to egalitarianism. Minds is clearly oriented towards professional content producers and advertisers, which reflects in their tokens being redeemable for guaranteed views. The platform directly relates the size of its token pool to the total expressed affect – in tokens but also in monetary value given their fixed price in USD. Contrarily, Steem issues a fixed number of tokens, whose exchange value, if discounted by speculative fluctuation, should reflect its capacity as “affective apparatus of capture” (Cockayne 2016:2). HyperSpace’s token pool size depends on both number of users and exchange value of tokens, which remains underdetermined.
In terms of governance, Steem’s quality as public, permissionless blockchain already shows in its general openness towards external developers. While Minds and HyperSpace predetermine the governance functions made available to their users and only decentralize decisions over their configuration and area of application, Steem leaves it to users to create all kinds of smart contracts, dApps and frontends and merely seeks to provide the foundational infrastructure for an emerging CSN. Its approach to enable users to develop unforeseen social media applications resembles second order prosumption that entails significantly more degrees of freedom than its competitors. The early stage of the platforms compared here does not allow detailed analysis of the power distribution each envisaged governance mechanism entails. However, noticeable tendencies are that Minds focuses on two-sided mechanisms while leaving aside group and community governance; HyperSpace focuses on group governance ultimately directed towards revenue distribution; and Steem is inclined towards equating capital with stake in decision-making. Each platform represents a distinct take on blockchain-based social media platforms that comes with up- and downsides: Minds as continuation of conventional Web 2.0 platforms enhances the possibilities of content producers to attract attention, but lacks attention to (groups of) consumers; Steem as anarcho-capitalist environment harnesses the productive potential of creative chaos but disproportionately empowers capital and suffers from a tragedy of the commons; and HyperSpace realigns incentives to gear interaction towards a collaborative outcome but lacks a business model giving its token value. The projects under scrutiny could easily suffer the fate of many other early attempts to build blockchain-based platforms – face unforeseen technical hurdles, run out of funding, be rejected by users, or become victim to criminal activity. The open-source infrastructure they build and the solutions they design for distributed governance may nevertheless become crucial elements of a new digital commons. This analysis has shown that outcomes for users in terms of autonomy, power and revenue distribution can vary significantly depending on discrete decisions over the design and alignment of tokenomics, governance and incentive mechanisms. It has also highlighted the continuous dependency on advertisements and other campaigns seeking attention and affect as sources of external demand for tokenized services. Interestingly, none of the platforms addresses this dependency on user attention by directly rewarding users for their audience labour. A tentative explanation that remains to be systematically scrutinized may be that user affect towards advertisements as economic driver of social media platforms itself must remain in the dark precisely because of its central role: While sociality as side product may be engineered and rationalized through incentive systems, genuine affect towards advertisements is the economically relevant product and as such may not be adulterated."