BRIEF SUMMARY OF HIELSCHER ET AL. (Reference Hielscher, Everding and Pies2022)

Hielscher et al. (Reference Hielscher, Everding and Pies2022) is built around the question of legitimacy and conditions for the legitimacy of private rule-setting, as observed in the sharing economy (see Reference Hielscher, Everding and Pies2022: 411). Thus, before critically commenting on this idea, let us show how the article argues that the private rule-setting of platforms such as Uber and Lyft is indeed legitimate.

From the very beginning, sharing platforms are mainly described as private actors that regulate and govern “a web of markets in which individuals use various forms of compensation to transact the redistribution of and access to resources” (Hielscher et al., Reference Hielscher, Everding and Pies2022: 404; Mair & Reischauer, Reference Mair and Reischauer2017: 12). On the contrary, it is just marginally mentioned how they often substantially act as traditional companies hiring their workforce and imposing strict rules on their clientele (Frenken & Fuenfschilling, Reference Frenken and Fuenfschilling2021). Consequently, for this “quasi-public” (Hielscher et al., Reference Hielscher, Everding and Pies2022: 404) function of market-matching systems and regulators of the consequent transactions emerging, the article considers it licit to apply James Buchanan’s (Reference Buchanan1975) constitutional contractarianism legitimacy test. That is, since these institutions allow stakeholders to gain some kinds of Pareto improvements by creating new markets, it is considered implicit that they could emerge from a hypothetical agreement among them and should be hence assessed as legitimate rule-setters.

Therefore, by applying a Buchanan-like contractarian framework (see also Hielscher, Beckmann, & Pies, Reference Hielscher, Beckmann and Pies2014), the article compares the welfare level enjoyed by sharing stakeholders in the previous status quo (not a hypothetical state of nature, but the institutional setting existing before the advent of commercial sharing platforms) and the level attained after certain rules established by sharing platforms are agreed upon. In this way, it suggests an equation between these rules and the constitutional and post-constitutional rules that in Buchanan’s constitutional political economy characterize a legitimate state (see Hielscher et al., Reference Hielscher, Everding and Pies2022: 409). This may be seen as a reckless innovation in applying the constitutional contract apparatus. To be sure, we have instead nothing to object to an extension of the method from state institutions to the domain of corporate law.Footnote ¹

However, the article almost takes for granted that sharing platforms’ rule-setting legitimacy is satisfactorily achieved with any Pareto improvement that stakeholders may gain because of the new regulation. Hence, it can quickly move from the analysis of the conditions for legitimacy to programmatic injunctions about how to improve this feature (see Reference Hielscher, Everding and Pies2022: 418). Namely, the article claims that, if sharing platforms respect some minimum conditions (see Reference Hielscher, Everding and Pies2022: 414), that is, guaranteeing exit and voice options to sharing partners (Hirschman, Reference Hirschman1970) while staying within the broader constitutional limits imposed by the basic structure of society (limits that can be then strengthened through additional “post-constitutional self-commitments”), they have the right to take upon themselves novel “ordo-responsibilities” (see also Pies, Hielscher, & Beckmann, Reference Pies, Hielscher and Beckmann2009) and exercise coercive power over other users. Therefore, they can unilaterally change the “rules of the game” for sharing markets, establish a “platform constitution” and, by implementing “constitutional and post-constitutional commitment services” such as Uber’s “deactivation policy” or Uber’s leasing program, further improve the game payoffs for all the other stakeholders (Hielscher et al., Reference Hielscher, Everding and Pies2022: 415–18).

The reason for this normative position is that the article does not interpret the institutional setting characterizing the beginnings of the sharing economy, before the birth of private platforms, as a system of emerging social norms in an equilibrium selection process and open to different possible equilibrium outcomes and juridical concretizations. On the contrary, it argues that that time was a sort of Hobbesian “state of nature” suffering from an “institutional deficit” (Reference Hielscher, Everding and Pies2022: 411), that is, a situation in which the lacking distribution of property rights over the Internet and the high transaction costs of finding reliable sharing partners, building reciprocal trust, and managing potential risks determined violent conflicts, unsecured sharing rights, and, hence, inefficient utilization of resources. Consequently, without the effort of private platforms to bridge that gap through constitutional and post-constitutional rules (see Reference Hielscher, Everding and Pies2022: 412), it would have been impossible to guarantee mutual betterment to all affected stakeholders. All of them would have been losing the opportunities offered by new technologies. Thus, whatever Pareto improvement for sharing stakeholders makes platforms’ regulation legitimate.

Accordingly, Hielscher et al. (Reference Hielscher, Everding and Pies2022) also adds a final claim that the newly established sharing constitutions must respect, answering its third fundamental question (see Reference Hielscher, Everding and Pies2022: 406). Namely, to be considered completely legitimate, sharing platforms necessitate conforming to an external “enabling institutional environment” in society established by public and civil actors to assist them in implementing their private regulatory framework (Reference Hielscher, Everding and Pies2022: 422). This enabling environment is described as a sort of “meta-constitution for platform constitutions.” However, in the article’s perspective, the meta-constitution is not intended as the higher level of the constitutional contract that establishes the system of rights and responsibilities that sharing platforms must respect and substantiate at the post-constitutional level. It is just presented as a “second-order” regulatory framework aimed to enhance sharing stakeholders’ exit and voice options by indirectly incentivizing sharing platforms to freely assume some extra commitments (Reference Hielscher, Everding and Pies2022: 425).

PLAN OF THE REPLY

In our reply, we present the alternative thesis that, even though some suggestions for improving sharing platforms’ rule-setting capacity may be sound, Hielscher et al.’s (Reference Hielscher, Everding and Pies2022) starting point—namely, that the sharing constitution established by these platforms is legitimate—is based on four widely questionable points. In particular, i) as a matter of fact, its empirical reconstruction of the sharing economy is partial because it does not properly stress how digital platforms are, in many cases, “employers” that should be held responsible for the conditions of their “workers” and may become monopolistic actors damaging consumers and third parties in society through invasive surveillance practices and anti-competitive behavior. ii) As a matter of interpretation, its normative reconstruction of constitutional contractarianism is flawed since it does not take into account post-Buchanan developments in the theory—even regarding the application of the constitutional/post-constitutional approach to business ethics—that provided an answer to the question about the fair and rational terms of the agreement that would be achieved in a bargaining game on the constitutional/post-constitutional contracts. A problem that Buchanan left unsolved. Similarly, iii) the article does not provide any positive analysis of the proposed social contract stability and thus does not explain how it can gain actual compliance—another aspect that was tackled by some relevant contributions developing Buchanan’s theory. Finally, iv) the description of the status quo ante of sharing markets, on which the entire argument is based, is not consistent with the evolution of the system of social norms that characterized peer-to-peer digital infrastructures before the advent of the sharing economy giants shifted the equilibrium selection path towards “neo-feudal” private appropriation of the Internet. Consequently, we conclude that the answer to the legitimacy question surrounding the unilateral rule-setting function of private platforms must be negative. However, the commentary also argues that the path conducive to the current situation of sharing markets was not the only possible one. Quite on the contrary, it shows that the cooperative form of platform governance would offer a solution to all the weaknesses of Hielscher et al. (Reference Hielscher, Everding and Pies2022).

THE BARGAINING PROBLEM: A UNILATERALLY DECIDED PARETO IMPROVEMENT DOES NOT AMOUNT TO A SOCIAL CONTRACT

As anticipated, empirically, one of the main problems of the article is that it primarily focuses on the idea that sharing platforms are “rule-makers” (Hielscher et al., Reference Hielscher, Everding and Pies2022: 404) of new markets, which they create through the matching of providers and users of goods and services thanks to digital means. In this way, however, the article seems to make two important omissions. On the one hand, it seems to conceal what has been called the “gig economy” (Woodcock & Graham, Reference Woodcock and Graham2020), that is, a highly significant sector within the sharing economy where hierarchical forms of business organizations and extractive practices—such as precarious working conditions, employment misclassification, and risk outsourcing (Bieber & Moggia, Reference Bieber and Moggia2021; De Stefano, Reference De Stefano2016)—are widely documented. In our opinion, the gig economy, an organizational model that de facto hides authority relations behind occasional market relations performed by supposed independent contractors, cannot be forgotten in a fair analysis of the phenomenon. This is especially true since platforms such as Uber and Lyft, always mentioned among gig platforms, are explicitly cited as examples.Footnote ² On the other hand, the article underestimates the increasing perplexities about users’ data extraction, algorithm surveillance, and behavioral manipulation practices of sharing platforms (Johnson & Acemoglu, Reference Johnson and Acemoglu2023; Marciano, Nicita, & Ramello, Reference Marciano, Nicita and Ramello2020; Martin, Reference Martin2015; West, Reference West2019; Zuboff, Reference Zuboff2015, Reference Zuboff2022) and the threat to fair competition connected with their monopolistic tendencies and their increasing privatization of essential infrastructures (Davis, Reference Davis2022; Khan, Reference Khan2017; Lindman, Makinen, & Kasanen, Reference Lindman, Makinen and Kasanen2023).

In a fully consistent way, but from a normative viewpoint, although a constitutional contractarian approach may be the right normative framework for addressing the tasks of regulating and governing sharing markets, we contend that the specific version of the social contract portrayed in the article risks to justify abuses of authority and thus cannot be accepted in a proper social contract perspective. In particular, we disagree with the thesis that whatever welfare improvement for sharing stakeholders reached through “Pareto-superior rules” (Hielscher et al., Reference Hielscher, Everding and Pies2022: 410) can legitimize the institution of a regulatory discipline for the sharing economy largely to the advantage of the party in the position of setting the rules. That is at least what would follow from the argument that a tiny (sometimes nearly null) utility improvement for most stakeholders is a sufficient condition for hypothetical agreement among them, even though the supposed agreement would result from a unilateral decision and lead to a sharply unfair outcome distribution.

Indeed, resting on the assumption that whatever Pareto improvement is equivalent to a mutual and rational agreement, Hielscher et al. (Reference Hielscher, Everding and Pies2022) transforms what would be supposed to be a bargaining game—wherein the parties converge on some bargaining equilibrium—into a noncooperative ultimatum game. This assumption clashes with the obvious economic modeling of the constitutional contract according to post-Buchanan developments in social contract theory (Binmore, Reference Binmore1997, Reference Binmore2005; Brock, Reference Brock1979; Gauthier, Reference Gauthier1986; Hampton, Reference Hampton1986) that see it as a bargaining game by which the parties reach a workable agreement within a fairly symmetrical set of feasible payoff distributions (on this point, see also Nash, Reference Nash1950). Coherently with these developments, in our account of the contractarian argument, the Pareto improvement condition is just a necessary but insufficient condition to characterize alone the bargaining solution.

To keep things simple, notice that, if the constitutional contract is seen as the solution of a Nash bargaining game with a symmetrical payoff space, this solution would coincide with an equal splitting of the surplus. But, even if we introduce asymmetries in the payoff space, the solution would always fall on the Pareto frontier point where the product of utilities is maximized so that the utilities’ marginal rate of substitution reflects the condition that the utility of one player cannot marginally decrease more than the marginal increase of the other player’s utility. This means that, given the convex shape of the bargaining space, it is not possible to largely reduce the utility of a party only in order to induce a small increase in the other’s utility, as would necessarily happen with a strongly unilateral bargaining solution. For these reasons, the article cannot provide any convincing normative argument for explaining why the most unbalanced possible concretization of the hypothetical agreement (like that emerging in an ultimatum game) could be ex ante preferred over the remaining points on the Pareto frontier of the bargaining space and be understood as a normative model corresponding to the idea of a constitutional contract.

On the contrary, given the possibility of selecting amongst many equally efficient points on the frontier, but entailing very different payoff distributions, a social contract seen as the rational solution of a bargaining game would rule out the option that a party is forced to accept just an infinitesimal part of the surplus. This is not only due to the normative constraints of impersonality and impartiality that are natural elements of the social contract way of reasoning but also follows from the typical condition implicit in any bargaining game—that is, that the result must be invariant with respect to the exchange of the players’ positions. However, that is actually a widespread experience for many low-skilled precarious gig workers who, being extremely dependent on their work and lacking any other alternative in the conventional labor market, often end up accepting whatever conditions platforms set (Schor, Attwood-Charles, Cansoy, Ladegaard, & Wengronowitz, Reference Schor, Attwood-Charles, Cansoy, Ladegaard and Wengronowitz2020).Footnote ³ At the same time, more subtle extractive practices and negative externalities are unilaterally imposed even on consumers and third parties in society due to the same platforms’ monetization on the accumulation of big data and monopolization of sharing markets by acting as gatekeepers and exploiting network effects.

In this sense, some scholars speak of “platform capitalism” (Srnicek, Reference Srnicek2017) or “neoliberalism on steroids” (Murillo, Buckland, & Val, Reference Murillo, Buckland and Val2017). For example, consider Uber’s so-called “deactivation policy.” The article, recalling the same arguments made by the platform, presents it as a way to prevent and sanction uncooperative drivers’ strategies in order to protect users (see Hielscher et al., Reference Hielscher, Everding and Pies2022: 419). However, deactivations may also be interpreted as traditional dismissals that the platform can implement without having to pay the costs of firing its employees. By accepting the unilateral role of rule-setters of sharing platforms, the article risks justifying, beyond the proposed reconstruction of the sharing economy, exactly these outcomes that are obviously unacceptable from a social contract perspective.

Apparently, this is a weakness that can be traced back to Buchanan’s theory. Indeed, because of the mutually destructive equilibrium of its state of nature, Buchanan (Reference Buchanan1975) claimed there is a mutual advantage in building a private property rights regime. Nevertheless, i) he did not prescribe a univocal theory of distribution since he lacked a proper bargaining theory, leaving many possible alternatives open. Furthermore, ii) he also suggested “salience” (Schelling, Reference Schelling1960) as a reasonable criterion to conjecture what agreement would be achieved by the constitutional contract, in particular by taking as a focal point the pre-social allocation of possession existing before the mutually destructive anarchical interaction started. Namely, the latter being a suboptimal equilibrium because of the reciprocal costs of investments in aggression and defense that the players must face in the state of nature, avoiding such costs and securing each player the initial possession would open the room for mutual advantage and guide to the identification of an agreement.Footnote ⁴ Therefore, even though there is no basis for inferring equality in Buchanan’s theory, there are also no reasons for deducing a seriously unbalanced social contract.

Accordingly, the aforementioned developments of contractarian theory in the following decades recognized how its completion would have required such a bargaining theory.Footnote ⁵ From this theory, it can be inferred, first, the (macro) constitutional agreement amongst all concerned individuals about the governance principles of economic transactions in general. Second, the following (micro) post-constitutional agreement about what ownership/organizational structure should be given to any economic institution (in this specific case a platform) to translate the agreed rights into practice (see Donaldson & Dunfee, Reference Donaldson and Dunfee1994, Reference Donaldson and Dunfee1995; Fia & Sacconi, Reference Fia and Sacconi2019; Sacconi, Reference Sacconi2006).

Moreover, to account for the self-sustainability that any social contract should satisfy for working as an effective normative foundation for a theory of economic institutions, the more sophisticated developments show that the constitutional contract should be modeled as an equilibrium selection device operating under a “veil of ignorance” (Binmore, Reference Binmore1997, Reference Binmore2005). In this way, the agreement boils down to a bargaining game over a symmetrical payoff space resulting from the intersection of the equilibrium space of the game played in the state of nature and its symmetrical translation operated by exchanging the players’ positions. That would identify the subset of equilibrium outcomes that satisfies the requisites of impersonality and impartiality. Then, by applying bargaining theory within this subset, it would be possible for the parties to select a unique and mutually acceptable agreement on the Pareto frontier. But notably, within such a subset—which is also symmetrical—the equilibrium solution needs to be the egalitarian Nash bargaining one (see Binmore, Reference Binmore2005: 175; see also Sacconi, Reference Sacconi, Sacconi, Blair, Freeman and Vercelli2011). Hence, as an implication, we may confirm that the parties in the contract would not be satisfied with whatever unilaterally selected Pareto improvement. They would instead require the respect of their equal rights obtained at the constitutional contract level to enter the post-constitutional agreement on the governance rules for sharing platforms.

THE COMPLIANCE PROBLEM: HOW NOT TO MAKE A NORMATIVE THEORY OF ECONOMIC INSTITUTIONS UTOPIAN

In this section, we focus on the problem of ex post compliance with regulations based on the social contract. Namely, the necessity of proving that the contract is self-sustainable not only against the test of ex ante acceptance based on impartial and impersonal reasons (justification) but also the ex post persistence test when stakeholders’ behavior reflects the complexity of their motivations (including self-interest and personal/collective attitudes). In this sense, we understand a normative theory as “non-utopian” or “realistically utopian” (Nagel, Reference Nagel1986; Rawls, Reference Rawls1999; see also the distinction between “internal” and “external” rationality in Gauthier, Reference Gauthier1986).

Accordingly, we will show that Hielscher et al. (Reference Hielscher, Everding and Pies2022) also fails to secure the stability of the recommended private regulation of the sharing economy because it underestimates the compliance problem and ignores its possible solutions. Namely, due to the underlying prisoner’s dilemma structure of the state of nature, the absence of any true external sanctioning authority and the contemporaneous lack of any stakeholders’ joint commitment resulting from a proper deliberative process, the contractors would have an individual incentive to defect ex post from the article’s proposed sharing constitution. Hence, missing the equilibrium property of the constitutional agreement (a property that we have introduced at the end of the previous section), the entire construction of the article would turn out to be unstable. This weakness is again a legacy deriving from Buchanan’s theory, and the same authors partially recognize it in the text (see Reference Hielscher, Everding and Pies2022: 429–30). Indeed, if the state of nature is a mutually destructive prisoner’s dilemma, as it is assumed, and even the compliance problem is a prisoner’s dilemma, the contractors’ psychology would be inherently self-interested and the only equilibrium solution (in dominant strategies) would be defection. This point was clearly seen by those who worked within the social contract tradition through the lens of game theory (Binmore, Reference Binmore1997, Reference Binmore2005; Gauthier, Reference Gauthier1986; Hampton, Reference Hampton1986; Skyrms, Reference Skyrms2003).

Therefore, lacking the equilibrium property, the platform constitution that Hielscher et al. (Reference Hielscher, Everding and Pies2022) propose would not have the strength required for becoming an equilibrium institution in Masahiko Aoki’s sense of the term—that is, “a self-sustaining system of shared beliefs about a salient way in which the game is repeatedly played” (Aoki, Reference Aoki2001:10)—and be capable of bridging the governance gap of sharing markets. Note that Aoki’s definition of institution is per se just a descriptive concept stressing that the stability of an institution requires a regularity of behavior in which the players’ actions are reciprocally optimal strategies and are sustained by a summary and shared mental representation of the equilibrium itself. However, such an equilibrium institution can only be achieved dynamically through an equilibrium selection process and normative mental models play an essential role in identifying the initial conditions of this process. In particular, the social contract reasoning can trigger the process by putting the players’ attempts of outguessing the reciprocal strategies on the path along which mutual predictions will converge on a mental representation of the game solution having the distributive property of a fair bargaining solution (see, on this point, Binmore, Reference Binmore2005; Sacconi, Reference Sacconi and Grandori2013). In this way, the social contract legitimizes the system of shared beliefs and supports the achievement of an equilibrium institution.

As we further argue in the next section, that is the reason why the instability of Hielscher et al.’s (Reference Hielscher, Everding and Pies2022) proposed sharing constitution is empirically proved by recent court decisions and growing supranational, national, and local rules enacted in many countries for regulating and constraining the activities of private platforms (from antitrust to privacy, tort, and labor law). Namely, these legal facts are exemplifications of normative mental models shared in society that affect the equilibrium selection process by reminding the players of the terms of a fair social contract that do not confirm the unilateral outcome based on private rule-setting (see Basu, Reference Basu2018 on the role of the law as the focal point of institutional equilibrium selection processes).

From what we know, within the theory of the social contract based on some version of economic modeling (so following the line started by Buchanan), the compliance problem could have been faced instead by Hielscher et al. (Reference Hielscher, Everding and Pies2022) in two different (but not incompatible) ways. The first one is the solution of Binmore presented at the end of the previous section, that is, formalizing the issue of self-sustainability of the social contract as an ex ante equilibrium selection problem with a multiplicity of possible equilibria played behind the veil (see Reference Binmore2005: 4). But in this case, as we have seen, admitting that the social contract can be an equilibrium, it can only be egalitarian. This result clashes with the institutional arrangement defended by the article once we realize that it includes a sector like the gig economy and justifies sharing platforms’ algorithmic surveillance and monopolistic tendencies. In fact, these equilibria would never be included in the symmetrical subset of the ex ante acceptable solutions and selected as a fair social contract.Footnote ⁶

The second way out of the compliance trap is to hypothesize that the fair agreement on an ethical norm of behavior may activate endogenously a variation in beliefs and preferences, so that, under the condition of reciprocal expectations of conformity, the participants in the ex ante agreement also prefer to comply ex post with the principles of the agreement itself, that is, develop conformity preferences. These preferences would make the social contract stable even if compliance would not have been the participants’ preferred behavior without the agreement. The topic of social preferences based on mutual expectations is not new in behavioral game theory (Rabin, Reference Rabin1993), the theory of social norms (Bicchieri, Reference Bicchieri2005), public choice, and business ethics. Among the possible options, the one we consider here is based on the link between pro-social individual preferences and the impartial agreement on the constitution of economic organizations (Grimalda & Sacconi, Reference Grimalda and Sacconi2005; Sacconi Reference Sacconi2007; Sacconi & Grimalda, Reference Sacconi, Grimalda, Bruni and Porta2007), a connection that has already been tested even through laboratory experiments (Degli Antoni, Faillo, Francés-Gómez, & Sacconi, Reference Degli Antoni, Faillo, Francés-Gómez and Sacconi2022; Faillo, Ottone, & Sacconi, Reference Faillo, Ottone and Sacconi2015; Sacconi & Faillo, Reference Sacconi and Faillo2010).Footnote ⁷

Actually, “conformity preferences” is just another term (with the addition of a game theoretical model) for what Rawls calls the “sense of justice” (Rawls, Reference Rawls1971, Reference Rawls and Kelly2001; Sacconi & Faillo, Reference Sacconi and Faillo2010). Indeed, for Rawls, the “sense of justice” is exactly an attitude of compliance with impartially agreed-upon principles based on the reciprocal expectation of conformity—these principles not being prima facie equilibria. Game theoretically, it corresponds to the activation of preferences that ex post modify the payoff structure of the compliance game so that the resulting “psychological game” (a game in which players are characterized not just by material payoffs but also by “psychological payoffs” depending on their mutual beliefs about conformity) offers equilibrium points to the players’ interaction that could not exist in the original game only based on material payoffs.Footnote ⁸ Therefore, activation of conformity preferences (or the sense of justice) would incentivize the parties endogenously to conform ex post with what they have jointly accepted ex ante, even if this, individually, would not be convenient.

Finally, note that this model, like most social preference theories, also attaches additional motivational weight to strategic choices that not only respect the ex ante agreement but deny other parties the possibility of deviating from it by intrinsically remunerating sanctioning. In our case, this would induce psychological payoffs for rejecting cooperation with a company that only minimally offers its stakeholders the benefits of their cooperation while appropriating much of the surplus. The recent forms of collective action that have characterized the sharing economy in the last few years, such as gig workers’ strikes and consumers’ boycotts, point exactly in this direction (Bunders, Reference Bunders, Meijerink, Jansen and Daskalova2021; Lehdonvirta, Reference Lehdonvirta2022; Woodcock & Graham, Reference Woodcock and Graham2020).

THE INSTITUTIONAL PROBLEM: THE INTERNET OF THE BEGINNINGS WAS NOT A RES NULLIUS BUT A SHARED INFRASTRUCTURE

Since Hielscher et al.’s (Reference Hielscher, Everding and Pies2022) version of the social contract is founded on a Hobbesian “state of nature” view of the Internet’s status quo ante, the article can argue that the unilateral rule-setting function of private platforms solved this inefficiency for the first time (see Reference Hielscher, Everding and Pies2022: 412). Hence, admitting that the previous critical points were not stringent enough, this interpretation would still justify the emergence of corporate governance institutions as the outcome of an implicit social contract putting an end to that mutually destructive state of nature. Nevertheless, we also suggest that the proper framing of the status quo is different. Namely, the Internet of the beginnings has been more appropriately framed as a shared infrastructure characterized by nonexcludability and only partial rivalry (Benkler, Reference Benkler2006; Frischmann, Reference Frischmann2005, Reference Frischmann2012; Fuster Morell, Reference Fuster Morell, Frischmann, Strandburg and Madison2014; Hess, Reference Hess2008; Kostakis & Bauwens, Reference Kostakis and Bauwens2014; Schultze & Whitt, Reference Schultze, Whitt, Bauer and Latzer2016).

In particular, Frischmann (Reference Frischmann2012: 3) describes shared infrastructures as a set of resources “made by humans for public consumption” and defined in terms of how they create value for their users, that is, through free downstream productive activities that generate positive spillovers. Therefore, in accordance with these features and because of its nonexcludable and (partially) rival character—two points that rule out both the possibility of centrally controlling and partitioning it without a significant loss of value—the shared infrastructure of the Internet boils down to a subspecies of Ostrom’s (Reference Ostrom1990) common-pool resources that can be governed, at least to some extent, by an evolving apparatus of collectively decided and self-sustaining rules. More specifically, Frischmann (Reference Frischmann2012: 320) qualifies the Internet as a complex and multilayered conglomeration of different infrastructure resources intertwined with each other. Accordingly, at the macro level, there is a congestible physical infrastructure—characterized by an interconnection of telecommunication, cable, server, and satellite networks—and a nonexcludable logical infrastructure, that is, the standards and protocols that facilitate the transmission of data across networks. Then, there are different applications running on this layer, such as social media platforms, online marketplaces, sharing platforms, and so on. Finally, there are the information, communication, content, and services that the users exchange through platforms and the social relations they create in doing that.

Coherently with this way of framing the Internet, a self-sustaining and collaborative alternative system of social norms and property rights, represented by still-existing and successful experiences such as the FLOSS (Free/Libre/Open Source Software) and the Wikipedia encyclopedia, was effectively developing at the outset for governing this common pool resource as a commons. According to the first enthusiastic reactions that it produced related to the promise of disintermediating and democratizing economic exchanges of goods and services among peers—ideally in the form of idle assets, time, or skills (see Botsman & Rogers, Reference Botsman and Rogers2010)—we can think of the early sharing economy as part of this narration. However, at least for the sharing economy’s sector, these emerging norms were not yet in a state of stable equilibrium reinforced by mutual expectations and shared mental models supporting them (Denzau & North, Reference Denzau and North1994) and were not concretized in any juridically recognized institution. On the contrary, they were just one of the different possible frames that were competing for saliency, among the traditional ones of private ownership—sometimes specified as club ownership—and state control. Thus, the early experiments of sharing platforms were still replaceable through an exogenous change of the initial conditions, determining a shift in the focal point of the equilibrium selection dynamic (recall Aoki’s definition of institutions).

In our interpretation, such a shift was provided by the venture-capital-backed big techs such as Uber and Lyft that, profiting from the existing legal vacuum, “enclosed” the Internet (or a portion of it) through the creation of capitalistic and legalistic private monopolies over a previously shared and open-source asset (Lehdonvirta, Reference Lehdonvirta2022; Muldoon, Reference Muldoon2022). Namely, the entry of several venture capitalists on the board of directors of the leading sharing start-ups determined the triumph of the for-profit mentality and shifted the equilibrium selection path towards the de facto reaffirmation of the hierarchical integration and economies of scale models typical of traditional capitalist companies (Frenken & Fuenfschilling, Reference Frenken and Fuenfschilling2021; Srnicek, Reference Srnicek2017). This is also the thesis developed by Pistor (Reference Pistor2019) about the capacity of private lawyers—informed by neoliberal culture—in designing new contracts and new property rights regimes over resources still not completely regulated and formerly accessible as shared resources and then mobilizing strategic and cognitive assets to impose and defend their private/exclusive appropriation.

Accordingly, someone has sustained that, behind the whole rhetoric of the sharing economy, “business as usual” soon reemerged and the same term “sharing economy” is now used only as a marketing expedient to cover outcomes that have nothing to do with the concept of sharing (Ravenelle, Reference Ravenelle2017). The story, however, did not end there, and many legal facts show that not even the institutionalization process of the self-regulated capitalist governance of the sharing economy has reached a stable state of equilibrium. On the contrary, new regulations and court decisions seeking to constrain the activities of sharing platforms, such as the 2017 sentence of the European Court of Justice which first framed Uber as a transport company instead of a technological market-matching system (European Court of Justice, 2017), are the proof that their unilateral authority is not accepted by all the affected stakeholders (who, from a social contracts viewpoint, would prefer a more equitable point on the Pareto frontier). Ironically, the article considers these legal facts as dysfunctional regulatory acts that only reduce stakeholders’ opportunities for mutual betterment (see Hielscher et al., Reference Hielscher, Everding and Pies2022: 424) rather than evidence that the regulatory system that emerged through private appropriation is unstable because of its inherent unfairness and the availability of fairer institutional arrangements along the frontier.

Consequently, from the lens of a different framing of the status quo, the juridical structure characterizing the governance as a commons (Ostrom, Reference Ostrom1990) could not be the problem but rather the solution to an only apparent institutional void and capable of avoiding the current concentrations of power with dubious legitimacy that characterize sharing markets. Indeed, instead of creating an unbalanced and radically unstable outcome, governing platforms as commons (and not as a state, club, or private good) could effectively produce a self-sustainable and equitable Pareto improvement, determining positive spillovers for all the sharing stakeholders coherently with the idea of the “comedy of the commons” (Frischmann, Reference Frischmann2005, Reference Frischmann2012).Footnote ⁹

Now, a peculiarity of sharing platforms is that they are organizations where the transaction of goods and services, even if enabled online by large tech companies, often takes place offline between people who can physically meet (think, for instance, to the ride-hailing sector) and who are in principle oriented by some kinds of pro-social motivations (Schor, Reference Schor2014). For this reason, the most consistent experience with the proposed alternative way of looking at the Internet is arguably the organizational model known as “platform cooperativism” (Nicoli & Paltrinieri, Reference Nicoli and Paltrinieri2019; Scholz & Schneider, Reference Scholz and Schneider2017). In fact, the idea of platform cooperatives consists of an attempt of “cloning or creatively altering the technological heart of the sharing economy” (Scholz, Reference Scholz2017: 174) and implementing the traditional cooperative principles of democratic governance and shared ownership by involving the relevant stakeholders in the property and control structures of digital platforms (see also Foramitti, Varvarousis, & Kallis, Reference Foramitti, Varvarousis and Kallis2020; Schneider, Reference Schneider2018).

Hence, on the one hand, platform cooperatives would respect the definition of infrastructure given above and avoid the private appropriation of the positive spillovers produced by their users, while the possibility of sharing the technology within a network of federated local cooperatives offers an alternative growth strategy to the building of big monopolistic giants (see Kostakis & Bauwens, Reference Kostakis and Bauwens2014 on the “global commons” scenario; Mannan & Schneider, Reference Mannan and Schneider2021; Scholz, Reference Scholz2023: 59). On the other hand, the potential criticism regarding their feasibility related to the difficulty of making possible continuous interaction between dispersed individuals (Bunders, Arets, Frenken, & De Moor, Reference Bunders, Arets, Frenken and De Moor2022) could be solved thanks to the peculiar features of sharing markets themselves—that is, usually, local scale of service provision and fairly homogenous stakeholders (at least locally). Indeed, following the governance theory, these features could guarantee the viability of the platform cooperative solution by fostering mutual trust (Spear, Reference Spear2000), engendering motivations based on ideological commitments (Rose-Ackerman, Reference Rose-Ackerman1996), and reducing transaction costs involved with bargaining (Hansmann, Reference Hansmann1996). At the same time, the intermediation of digital technology potentially makes even easier their success as compared to traditional cooperatives by facilitating organizational efforts and the detection of free riders (Belloc, Reference Belloc2019).

DISCUSSION: HOW PLATFORM COOPERATIVES CAN SOLVE THE IDENTIFIED WEAKNESSES AND ADDRESS CRITICISMS

Drawing the appropriate conclusion from the more nuanced constitutional contractarian theory presented in this commentary, the purpose of this section is to discuss how the institutional alternative of platform cooperatives can address all the problems of Hielscher et al. (Reference Hielscher, Everding and Pies2022). In short, platform cooperatives can be seen as i) the most legitimate organizational structure for substantiating the regulation of sharing markets hypothetically bargained at the constitutional level in line with the effective reality of the sharing economy, ii) providing incentives and motivations consistent with its endogenous self-sustainability, and iii) proving to be more coherent with the idea of a Pareto improvement from an ex ante status quo representing the nature of the Internet as shared infrastructure.

Let’s start by hypothetically involving all the relevant stakeholders (see Freeman, Harrison, Wicks, Parmar, & De Colle, Reference Freeman, Harrison, Wicks, Parmar and De Colle2010) in the impersonal and impartial collective ex ante bargaining of the platform constitution and let’s assume that their agreement concludes that platforms have to be treated as commons. Their constitutive social contract would require recognizing to all the ideally contracting stakeholders’ freedom of access and participatory rights, the limited list of property rights compatible with the notion of collective ownership of a common pool resource (see Schlager & Ostrom, Reference Schlager and Ostrom1992), and would give them the entitlements enabling the establishment of the most effective governance structure to realize these rights.

Coherently with the idea of an emerging social norm legally actualizable in different possible ways by a local social contract, many options are open for detailing this governance structure. Among these options, multi-stakeholder platform cooperatives would be the most natural candidate.Footnote ¹⁰ Indeed, multi-stakeholder cooperative arrangements would symmetrically satisfy the quest for fairness created by the social contract procedure and hence generate mutual expectations for conformity with their democratic rules. Conformity preferences (or the sense of justice) would be thus elicited and stabilize these arrangements ex post. And finally, from an institutionalist point of view, they would also be coherent with the original representation of the Internet as a multilayered common pool resource, thanks to their inherent feature of giving a say in the governance at different levels to all their stakeholders (see again Mannan & Schneider, Reference Mannan and Schneider2021).Footnote ¹¹

However, assuming Hielscher et al.’s (Reference Hielscher, Everding and Pies2022) perspective, companies such as The Drivers Cooperative,Footnote ¹² a New York City-based ride-hailing platform cooperative with a multi-stakeholder inspiration, could be still criticized for their apparent incapacity to guarantee the same win-win-win(-win) opportunities to platform owners, directly involved stakeholders (both consumers and service providers), and indirectly involved third parties as compared to their commercial competitors. Indeed, after having stated the equal legitimacy of commercial platforms and platform cooperatives from a social contract perspective, Hielscher et al. (Reference Hielscher, Everding and Pies2022) claim that commercial platforms are in the end preferable. Namely, according to the article, commercial platforms offer stronger exit options to sharing partners and thus a superior possibility of mutual betterment because they are more ready to face market competition and capable of continuously innovating themselves to match the evolution of their users’ expectations (see Reference Hielscher, Everding and Pies2022: 428–29).

We believe that even this point is misleading because, in the long run, it is difficult to sustain that the current business model of the sharing economy is going to produce the promised win-win-win(-win) outcomes and consequently offer a true competitive advantage to its different stakeholders (see Reference Hielscher, Everding and Pies2022: 416). First, commercial platforms are not currently creating any real profit for their investors and owners. On the contrary, their business model—that is, “growth before profits”—seems to consist only of reinvesting all their earnings in disrupting competitors with the aim of a future dividend distribution that has yet to be achieved. Uber is again an emblematic example. In the meantime, venture capitalists are instead losing and not gaining money (Isaac, Reference Isaac2019). Secondly, if we intend the directly involved stakeholders as consumers, it is logical to imagine that, if sharing platforms finally do achieve their aspired monopolistic position, this will also lead to a rise in prices and a reduction of benefits for consumers themselves (Khan, Reference Khan2017). Moreover, some scholars have even stressed how twenty-first-century consumers are becoming “prosumers” (Ritzer, Reference Ritzer2010), invisible workers who, while consuming, produce information and data that will be used by unaccountable platforms to influence their choices and create lock-in effects without paying them for those data (Marciano et al., Reference Marciano, Nicita and Ramello2020; Martin, Reference Martin2015; Zuboff, Reference Zuboff2015). If, instead, we intend the directly involved stakeholders as service providers, that is, gig workers, we have already assessed how commercial platforms are reintroducing hierarchical organizational forms typical of traditional capitalist enterprises that often deprive their working conditions (Frenken & Fuenfschilling, Reference Frenken and Fuenfschilling2021; Srnicek, Reference Srnicek2017). Finally, they also seem to produce negative externalities for third parties in society by accelerating the transition to a labor market characterized by precarious jobs and making the whole economy less sustainable, innovative, and inclusive in the long run while disenfranchising people from the control of essential infrastructures of their daily lives (Crouch, Reference Crouch2019; Davis, Reference Davis2022; Johnson & Acemoglu, Reference Johnson and Acemoglu2023; Khan, Reference Khan2017; Lindman et al., Reference Lindman, Makinen and Kasanen2023; West, Reference West2019; Zuboff, Reference Zuboff2022).Footnote ¹³

Oppositely, a cooperative platform such as The Drivers Cooperative takes a smaller commission than Lyft or Uber, pays its workers a guaranteed hourly wage of $30/hr (almost twice NYC’s minimum wage), aims to redistribute profits among members in the form of dividends, lets them participate in the governance of the company (even for what regards data and algorithmic management), shares relevant information with all the stakeholders, addresses traditional concerns such as predatory loan rates and surprise deactivations, and fights precariousness. Moreover, it also charges clients a lower fare, wants to have a positive social impact by enhancing service quality for people with disabilities (being a leading provider in New York City in the niche of accessible social mobility thanks to specific subcontractor agreements with paratransit brokers operating for the Metropolitan Transportation Authority), and is finally planning to share its driver-owned platform, created from scratch thanks to a partnership with Google, with federated platform co-ops from all over the world in order to scale as its corporate competitors—starting from another ride-hailing platform that will be launched in Denver in 2024 (Conger, Reference Conger2021; The Drivers Cooperative, 2023).Footnote ¹⁴