Governance, technology, and the limits of digital solidarity economies: A South Korean case study

1. Introduction

South Korea’s current push for people-centred Fourth Industrial Revolution (4IR) platforms unfolds at the crossroads of two powerful yet often conflicting legacies: a developmental state long focused on technological catch-up (Kim, 1997; Minns, 2001) and a Social Solidarity Economy (SSE) lineage grounded in mutual-aid practices such as dure, gye, and hyangyak (Jung & Rösner, 2012; Ji, 2018). Provincial governments now sponsor ride-hailing, food-delivery, and mobility apps under the banner of public ownership, promising to blunt private monopolies and revive local economies (Kim & Choi, 2019; PCFIR, 2018; Seo, 2022). Unpacking whether these experiments genuinely democratise digital infrastructure – or merely re-package enclosure – has pressing consequences for both platform-economy scholarship and public policy.

Policy white papers situate such initiatives within a three-pillar SSE agenda – urban regeneration, village-community building, and social-economy enterprise (Kim, 2023; Jin, 2024). Urban-commons studies likewise stress community self-governance and affordable space (Park et al., 2020; Paris et al., 2024). Yet few accounts trace how this discourse collides – or colludes – with bureaucratic planning cultures and private-procurement norms once digital projects leave the drawing board.

A survey of overlapping literatures underscores the stakes. Platform-capitalism research shows how commercial platforms convert data, labour, and logistics into rent (Srnicek, 2017; Birch & Ward, 2022), deepening precarity via algorithmic scheduling (De Stefano, 2016; Graham et al., 2025). Science-and-Technology Studies (STS) cautions that “smart” infrastructures may simply re-package enclosure (Paris et al., 2024), while Latin-American STS adds post-colonial critiques of platform roll-outs (Albornoz, 2023). Platform-co-operativism envisions democratically owned services (Scholz, 2016, 2023) and sometimes fares better on Fairwork ratings (Graham et al., 2025), though tensions between activism and commercialisation persist (Sandoval, 2020). Taken together, these strands rarely converge on the micro-institutional mechanics that translate national visions into day-to-day platform governance.

The concept of sociotechnical imaginaries helps bridge this gap, highlighting how national visions of progress guide technology policy (Jasanoff & Kim, 2015). South Korea’s dominant imaginary portrays science and technology as engines of rapid development, casting publics as “dutiful citizens” expected to support growth-first agendas (Song, 2003). Integrating this lens with political-economy debates exposes two persistent blind spots:

• Gap 1 – Governance rhetoric vs managerial reality. Scholarship richly documents how private platforms extract value through data monopolies and gig-labour regimes (Srnicek, 2017; Sadowski, 2020). Far less is known about state-affiliated platforms that promise democratic ownership. Even where cooperative rhetoric abounds, control often reverts to managerial elites, vendors, or legal vehicles insulated from public oversight (Hansmann, 1996; Scholz, 2016; Sandoval, 2020). How do instruments such as special-purpose companies or public-private contracts translate participatory slogans into genuine co-governance – or into performative participation that masks hierarchy?
• Gap 2 – Sociotechnical constraints on participation. Critiques of technosolutionism (Paris et al., 2024; Albornoz, 2023) seldom follow the trail of how sophisticated infrastructures – blockchain ledgers, AI routing engines – mediate data access, labour conditions, and collective agency, especially amid digital divides and legacy systems. Empirical evidence is needed on how architectural choices and uneven digital literacies co-produce new exclusions or openings for collective control.

Bridging these gaps prompts two questions:

• RQ 1. How are the sociotechnical imaginaries of convenience and social value articulated in a government platform proposal translated – or mistranslated – into concrete governance structures, ownership arrangements, and day-to-day practices?
• RQ 2. Despite public-ownership rhetoric, in what ways does the implemented platform reproduce rentier or extractive dynamics characteristic of corporate platform capitalism, and how do local stakeholders perceive and contest those dynamics?

By interrogating an anonymised “City P” mobility experiment – tracking its trajectory from the initial proposal, through a brief pilot phase, to suspension several years later – this study offers a unique window onto the frictions among national imaginaries, local politics, and the material affordances of digital infrastructure.

This article contributes by (i) extending Hansmann’s ownership calculus to public-sector platform cooperatives; (ii) operationalising STS constructs – black-box depth and participation performance – within a developmental-state context; and (iii) proposing design heuristics for policymakers who seek to align solidarity-economy rhetoric with enforceable governance rules.

Section 2 elaborates the theoretical framework; Section 3 details data and methods; Section 4 presents findings; and Section 5 discusses implications and concludes.

2.  Four lenses on digital solidarity platforms: Governance, ownership, solidarity, and sociotechnical imaginaries

This chapter assembles four complementary theoretical lenses and one integrating framework to interrogate a state‑sponsored mobility‑platform pilot. Section 2.1 revisits collaborative‑governance theory, pinpointing when facilitative mediation slips into merely symbolic consultation. Section 2.2 applies Hansmann’s enterprise‑ownership calculus to test which patron group can most efficiently exercise residual control over a “people‑centred” infrastructure. Section 2.3 excavates Korea’s pre‑modern mutual‑aid institutions – dure, gye, and hyangyak – to historicise contemporary platform‑co‑operative claims. Section 2.4 couples digital‑solidarity‑economy scholarship with the STS concept of sociotechnical imaginaries. Drawing inspiration from critiques of 'black-box' infrastructures (Paris et al., 2024), we operationalise two specific metrics for our analysis: “black‑box depth” and “participation performance”. Section 2.5 then weaves these strands into an analytic matrix that links vision, design, practice and emergent political‑economic dynamics, providing the benchmark against which our two research questions are assessed.

2.1 Collaborative governance: How governments broker consensus among unequal stakeholders

Collaborative governance denotes arrangements in which public agencies convene citizens, civil‑society organisations, and market actors in shared decision‑making forums rather than relying on adversarial bargaining or top‑down hierarchy. Ansell and Gash (2008) identify four interlocking success conditions: (i) a favourable starting context, (ii) transparent institutional design, (iii) facilitative leadership, and (iv) an iterative cycle of small wins. When aligned, collaboration can progress from symbolic consultation to genuine co‑production; when misaligned, participation becomes largely performative – public briefings replace negotiated authority while design and code remain centralised. By performative participation, we mean processes that ask people to attend meetings or trainings without giving them enforceable say over rules, budgets, or data. When decisions cannot be changed by those affected, participation looks inclusive but does not travel into practice. We treat this as a diagnostic sign of fragile collaboration. Public digital‑platform projects test these conditions severely. They mobilise at least four stakeholder blocs – local residents, incumbent mobility providers, technology vendors, and multiple levels of government – whose power and knowledge resources differ starkly (Chrislip & Larson, 1994). Early design choices in data architecture, pricing algorithms, or API accessibility can lock in governance outcomes long before deliberation occurs. Our analysis asks whether such safeguards existed – and where breakdowns emerged – when the City P platform travelled from proposal to implementation.

2.2 Enterprise‑ownership theory: Identifying the lowest‑cost controllers of public digital infrastructure

If collaborative governance describes how actors ought to bargain, enterprise‑ownership theory explains who ultimately controls and benefits. Hansmann (1996) argues that the most efficient ownership form minimises the combined market‑contracting costs faced by non‑owners and the governance costs borne by owners themselves. Applying his calculus to a government‑backed mobility‑as‑a‑service (MaaS) platform raises three questions. First, which patron group – drivers, resident lot‑owners, municipal agencies, or venture vendors – incurs the lowest sum of governance and contracting costs? Second, do instruments such as special‑purpose vehicles or data trusts actually transfer residual control to that low‑cost class, or do they entrench a higher‑cost surrogate, inviting managerial capture? Third, how do digital assets – sensor data, routing algorithms, blockchain ledgers – alter the cost matrix, potentially shifting optimal ownership from labour to code proprietors?

Recent civic‑tech scholarship re‑examines Hansmann in the platform era. Scholz (2016) shows how user‑ and worker‑owned platforms leverage software to lower classic governance costs (smart ballots, ledger transparency), while Sadowski (2020) warns that proprietary infrastructures permit new data‑rent extraction. Hence, we scrutinise not only share registers but also code repositories, data licences, and revenue rules, asking whether City P’s “community ownership” reduced or simply displaced the frictions Hansmann describes.

2.3  Korean mutual-aid institutions: dure, gye, and hyangyak

South-Korean policymakers routinely evoke three pre-modern mutual-aid forms – dure, gye, and hyangyak – to lend moral gravitas to contemporary Social-Solidarity-Economy (SSE) programmes (Ji, 2018; Jung & Rösner, 2012). Although now blurred together in official brochures, each institution arose from distinct material needs and encoded a different governance logic; tracing their trajectories into the present illuminates both the imaginative appeal and the practical limits of “people-centred” digital platforms such as the City P project.

Hyangyak (향약 or village compacts) constituted the most formal layer of village self-rule. Introduced by neo-Confucian magistrates in mid-Chosŏn, a hyangyak compact listed virtuous acts, prescribed fines for misconduct, and authorised public shaming boards to discipline free-riders. Its four classical principles – mutual encouragement of virtue, mutual regulation of faults, mutual decorum, and mutual aid in hardship – bound welfare to moral surveillance, producing what later statesmen would describe as “dutiful citizens” aligned with national goals (Ji, 2018). Modern consumer co-operatives echo this regimen when they publish audit tables and censure members who breach ethical-sourcing bylaws.

Dure (두레 or labour pooling), by contrast, emerged organically from agrarian labour demands. During planting and harvest, households contributed workdays to a rotating pool, shared meals at the host’s yard, and imposed social sanctions on shirkers. The system cut transaction costs, insured against crop failure, and embedded reciprocity in everyday toil (Choi, 2006). Its ethos resurfaced in the 1970s Saemaul Undong, when villagers paved roads or re-roofed houses in return for cement coupons – an explicit state re-packaging of the dure labour-swap logic.

Gye (계 or rotating credit) fulfilled community financial needs. Functioning as a rotating savings and credit association (ROSCA), members paid regular dues into a pot that was redistributed each cycle by lottery or auction, funding weddings, funerals, or entrepreneurial starts. Trust substituted for collateral, and default risk was managed through face-to-face accountability (Jung & Rösner, 2012). Post-liberation credit-union legislation scaled this mechanism nationally: today’s community credit cooperatives still rotate surplus back to patrons and publish payout schedules, retaining the gye spirit in formal guise (Ji ,2018).

These lineages stretch into the present SSE landscape. The National Agricultural Cooperative Federation (Nonghyup) returns purchase rebates – echoing gye dividends – to 2.2 million farmer-members each year. Credit unions (Shinhyup) manage 5 million accounts on the same rotating-benefit principle. Consumer co-ops adopt hyangyak-style moral compacts to police chemical-free supply chains, while neighbourhood social enterprises and worker co-ops invoke dure reciprocity to justify profit-sharing rules (Ji, 2018). Indeed, the Framework Act on Cooperatives provided the definitive legislative scaffolding for this movement, enabling thousands of new co-ops inspired by this “traditional mutual-aid culture” to form (2012).

Let's see what these traditions looked like in practice. A dure day meant neighbours rotated their labour to complete a harvest, with agreed sanctions for no‑shows. A gye meeting pooled small monthly dues and paid out the pot to one member each round, recorded in a notebook everyone could inspect. A hyangyak compact set out shared rules – what behaviour was expected, what fines applied – and posted them publicly. Translated into modern community enterprises, these logics suggest three design anchors: a rota for shared work, a ledger that any member can read, and a compact that gives residents clear rights and remedies.

Yet continuity is uneven. Thousands of post-2012 co-ops remain micro-scale, and few digital platforms achieve both commercial viability and democratic depth (Kim & Im, 2015). The rhetoric of dure-style solidarity or gye-based commons funds can therefore mask managerial capture when code repositories, tariff algorithms, or data ledgers stay locked behind vendor non-disclosure agreements. In the pages that follow, we examine how City P’s platform deployed these cultural signifiers and where, under the pressure of bureaucratic timetables and black-boxed infrastructure, performative participation supplanted substantive community control.

2.4 Digital solidarity economy and sociotechnical imaginary

Digital solidarity economy (DSE) research recasts online infrastructures as commons stewarded by the workers and users who rely on them. Anchored in critical political economy, the DSE agenda advances shared ownership not merely as a moral claim but as a structural antidote to assetisation, data rent, and algorithmic precarity. Early platform-co-operative cases show that, when technical sovereignty, patient capital, and supportive regulation converge, co-ops can secure fairer labour conditions and keep data value circulating locally (Scholz, 2016; Grohmann, 2023; OECD, 2023). Yet comparative evidence documents a recurrent slide toward “managerial recapture”: liquidity shortages, vendor lock-in, and opaque contracts quietly reinstate hierarchical control despite cooperative branding (Sandoval, 2020).

To explain this disjuncture between promise and practice, scholars invoke sociotechnical imaginaries – collective visions that tether technological advance to particular social orders (Jasanoff & Kim, 2015). These imaginaries act as performative blueprints: they steer architecture, allocate risk, and designate legitimate participants long before any code is written. Two blind spots persist. First, an alignment gap: emancipatory rhetoric often co-exists with data monopolies, assetisation tactics and legal clauses that reproduce extraction (Srnicek, 2017; Sandoval, 2020). Second, an infrastructural gap: deep “black-box” layers – from blockchain ledgers to AI dispatch engines – can neutralise formal voting rights by shielding critical parameters from democratic scrutiny (Paris et al., 2024).

Accordingly, this article operationalises three diagnostic variables – black-box depth, imaginary alignment, and participation performance – to trace how DSE ideals travel from vision to code to everyday governance (Paris et al., 2024). Treating imaginaries not as lofty slogans but as testable design hypotheses enables researchers, policymakers, and communities to pinpoint the junctures where solidarity crystallises into enforceable rules – or is relegated to branding. Understanding where that antidote succeeds or fails is essential for those building resilient post-platform digital commons, and it furnishes a rigorous yardstick for assessing future cooperative-platform experiments.

2.5 Integrative framework: Linking vision, design, practice, and political‑economic outcomes

Section 2.5 synthesises the four preceding theoretical perspectives into a single integrative framework. Rather than treating each lens in isolation, the framework sets them into a cross-tabulated matrix (presented in Table 1) that traces how ideas travel from vision to institutional design, into lived practice, and culminate in emergent political-economic outcomes. Collaborative governance, enterprise ownership, mutual-aid traditions, and STS imaginaries each articulate a distinctive horizon of what a “people-centred” platform should look like; the table juxtaposes these horizons to show how they prescribe different institutional architectures, produce varied forms of everyday practice, and yield contrasting trajectories of legitimacy, capture, or solidarity. This crossing of lenses highlights points of alignment and dissonance, furnishing both a diagnostic map of where solidaristic ambitions unravel and a benchmark against which to assess our research questions, present our findings and discussions sections.

3. Methodologies

3.1 Research design

We adopt a critical single-case design that follows the City P shared-mobility platform from proposal to suspension. The case is “instrumental” in Yin’s (2018) sense – selected not for its representativeness but because its full life-cycle lets us test how a state-led digital platform travels through the Vision → Design → Practice → Outcome chain set out in Section 2.5. Such an “embedded process‐tracing” logic privileges analytic over statistical generalisation, allowing theory refinement through thick description rather than sample size (Yin, 2018).

Two features make the case especially suited to our research questions. First, the platform explicitly claimed to operationalise SSE principles, giving us a real-world laboratory in which to observe whether solidarity rhetoric converts into enforceable rules. Second, the project compressed the entire platform life-cycle into four years, enabling within-case comparison across phases without cross-case noise. Throughout, we treat municipal planners, vendor firms, and resident groups as nested units whose interactions crystallise the political-economic outcomes we seek to explain.

Our empirical materials span four data types:

• Policy and planning documents (municipal proposals, steering-committee minutes, Special Purpose Company or SPC¹by-laws);
• Ethnographic fieldnotes (from two public sessions and a suspension forum);
• Local media and resident commentary (newspaper articles, council proceedings, social media threads);
• Interviews (n=11) with municipal staff, vendors, cooperative board members, and resident gig workers.

3.2 Case background: City P’s shared-mobility platform

City P is located within the commercial core of a mid-sized provincial municipality in South Korea. Once a vibrant local centre, the area experienced steady decline in the wake of industrial relocation and suburban sprawl. The old-town zone – composed of narrow lanes, ageing single-storey houses, and long-standing curb-parking habits – gradually came to symbolise both logistical dysfunction and socioeconomic stagnation. Years prior to the mobility project, residents had mobilised successfully to block a major redevelopment initiative. That collective experience left two defining legacies: strong intra-village kin networks and a deeply ingrained suspicion toward top-down urban interventions.

Around 2020, City P was selected as a demonstrator site for a Digital Social Solidarity Economy (DSSE) pilot funded through central government innovation programmes. Framed as a “people-centred” DSSE initiative, the programme envisioned a resident-owned cooperative – SPC-Coop P – operating shared parking and micro-mobility services. Identity and settlement would run on a city-wide pass integrated with Decentralised Identifiers (DID)/blockchain technology, with all benefits designed to be auditable and recycled locally. In official terms, municipal planners cast the city as a neutral facilitator, but in practice much of the design and implementation authority migrated to a private Service-Vendor Consortium, composed of software and hardware contractors headquartered outside the region.

The technical system was built on four layers: (1) smart cameras to detect parking violations, (2) a permissioned blockchain ledger to log every transaction, (3) a Mobility-as-a-Service (MaaS) app integrating parking, scooters, and micro-Electric Vehicles, and (4) a token-reward system that converted neighbourhood contributions (e.g., making driveways available during peak hours) into local shop credits. The SPC board, largely composed of long-time clan leaders and community representatives, was tasked with managing revenues and overseeing surplus allocation through a “Mobility Commons Fund.” Yet many board members later reported that they lacked access to real-time usage dashboards and had minimal say over algorithmic pricing decisions. In plain terms, the platform attempted to stitch together three functions – managed street parking, short-trip shared vehicles, and a rewards pool that circulated benefits to local shops – but failed to deliver the basic governance tools residents expected, such as transparent dashboards or binding approval rights over tariff changes.

A brief process timeline highlights this drift. Selection and early outreach (Year 1) were followed by the rapid painting of parking lines and a soft launch of the app. Short-term indicators looked positive – city slides reported fewer incoming cars, a stronger match between spaces and demand, and 21 local jobs created – but daily frictions mounted (confusing sign-ins, unclear rules). As lines appeared without clear agreements, objections and rumours surged. An independent conflict report recorded roughly 1,500 complaints at the peak (Conflict diagnosis report, 2020). Public briefings continued with little power to change rules; a suspension forum was then convened and operations halted.

By the time of suspension, the project’s original vision of community ownership and transparent, people-centred infrastructure had eroded. Residents described the platform as “externally driven” and “technically opaque.” What began as an ambitious demonstration of DSSE ideals devolved into a cautionary tale about the risks of platform governance without genuine participatory oversight.

3.3 Data analysis: Reflexive thematic coding

We employed reflexive thematic analysis (Braun & Clarke, 2012) to surface patterned meanings while remaining attuned to our positionality as analyst-participants. The procedure unfolded in five iterative steps:

1. Familiarisation. All transcripts and documents were read twice; analytic memos captured initial hunches.
2. Open coding. Segments expressing power, ownership, reciprocity, or technopolitics received inductive labels.
3. Framework mapping. Codes were clustered into a 16-cell matrix that crosses the four analytic lenses – Collaborative Governance, Enterprise Ownership, Solidarity & Mutual-Aid, STS/Participation Performance – with the four phases – Vision, Design, Practice, Outcome – proposed in Section 2.5.
4. Theme construction. Within each cell, we aggregated codes into higher-order themes (e.g., “tokenised reciprocity stalls at pay-out layer” in Solidarity/Practice).
5. Iterative refinement. Themes were reviewed against raw excerpts and challenged in peer debriefs until they accounted for deviant cases and inter-phase dynamics.

This matrix served as a visual audit trail linking data to theory, while the reflexive stance obliged us to document how our prior roles and normative commitments shaped interpretive choices. By focusing on what is said (content) and when it becomes consequential (sequence), the analysis avoids the prolix code-lists typical of template methods and keeps the spotlight on the integrative framework. Operational definitions used in coding. We read ‘black‑box depth’ as the number of core operating decisions that ordinary members cannot see or amend (for example, how prices are set, who can change rules, and who can view the logs). ‘Imaginary alignment’ is the fit between public promises (e.g., “community‑owned”, “local benefits”) and the actual allocation of rights in contracts, dashboards, and meetings. ‘Participation performance’ is the extent to which meetings and trainings lead to enforceable changes in rules or practice. These working definitions guided both our excerpt coding and the cross‑phase matrix in Section 4.

3.4 Credibility, ethics, and analytical value

Credibility was enhanced through source triangulation (documents, observation, interviews) and member checks in which key informants reviewed excerpt grids and timeline reconstructions. The lead author’s previous work in SSE policy provided deep access but also risked confirmatory bias; a reflective diary and external peer auditor mitigated this by flagging moments where advocacy leanings threatened analytic balance.

The study received Sungkyunkwan University IRB approval in 2024. Participants gave written or verbal consent, could opt out of recording, and vetted anonymised quotations. All organisational identifiers have been masked to protect confidentiality in a small-community context.

By tying a single-case narrative to a phase-sensitive, multi-lens matrix, the methodology does more than catalogue success or failure; it pinpoints where along the Vision → Outcome chain solidarity ambitions unravel and how technical design choices enable managerial capture. This scaffolding positions the subsequent findings to speak simultaneously to platform-governance theory and to practitioners searching for actionable design heuristics. Neither the municipal sponsors nor any external funder influenced the research questions, data access, analysis, or the decision to publish; all interpretations remain our own.

4. From promise to breakdown: Findings through four theoretical lenses

We structure our analysis through the four lenses established in our framework – collaborative governance, ownership & control, solidarity & mutual-aid, and STS/participation performance. Within each lens, we follow the Vision–Design–Practice–Outcome chain (see §2.5) to explain where and why the promised trajectory diverged from lived experience.

4.1 Collaborative governance: From inclusive vision to hollowed-out practice

Vision

City P’s official proposals promoted a facilitative, resident-led governance model, framing the project as a direct response to the community's history of top-down interventions. Officials pledged a process of staged briefings, capacity-building workshops, and a project council where residents would act as genuine co-planners and co-decision makers. The official storyline promised to leverage the visible pain point of curbside parking to transform a historically divided neighborhood into a community of “co-problem solvers.” This was a conscious attempt to manufacture the favorable starting conditions that collaborative governance theory deems essential for success (Internal project proposal, 2020).

Design

This inclusive vision translated into a series of process-oriented deliverables – manuals, training schedules, and meeting minutes – but crucially not into binding procedures or clear veto rules for residents. The project’s governing annexes specified the creation of workshops and monitoring tools but omitted the institutional thresholds that would grant residents substantive power. For example, there was no requirement that residents approve tariff changes or data-sharing agreements before implementation. This design flaw left the power imbalance between residents, the city, and vendors unaddressed. A technology vendor, P11, warned early that mobilisation would be fragile, noting, “This is an aging neighborhood… elders stay home, and many show little interest in these projects.” This observation hinted at deeper issues of digital literacy and accessibility that the project’s design failed to adequately address.

Practice

In reality, the project’s bureaucratic timeline meant that infrastructure development preceded the community co-decision process, thus creating conflict. Residents described how the sudden painting of parking lines on familiar streets sparked an intense backlash. “People turned against us because of the lines – ‘Why can’t I park in front of my house? Why pay now?’” recalled P3, a clan-village leader. This reaction revealed a fundamental failure to build trust before altering the physical environment. Even a resident champion, P2, admitted her own confusion: “At the beginning I didn’t even know what this platform was.” The rush to implement, characteristic of the state’s developmentalist culture, overrode the need for deep consensus. A city official acknowledged the procedural failure: “It was rushed; the education period was too short – we probably needed a year.”

Outcome

While meetings continued to be held, their credibility and influence waned over time. Instead of serving as forums for co-decision, they became sites for one-way information delivery from officials to residents. Eventually, boycotts and personalised rumors replaced deliberation, corroding the social fabric the project aimed to strengthen. “It got so bad that even a local restaurant refused to deliver food to our meeting,” recalled a resident organiser, a stark indicator of how social capital had been destroyed. Collaborative forms persisted on paper, but their substance had evaporated. Governance shifted from the ideal of “we decide together” to the reality of “we are informed together,” a clear slide from the promise of co-production into performative participation.

4.2 Ownership and control: Cooperative in name, extractive in design

Vision

Economically, SPC-Coop P was designed to localise value and ensure profits remained within the community. Official documents promised that 70–80% of any surplus would return directly to the cooperative, with the entire process guaranteed by DID/blockchain technology to ensure “trust through code (Internal project proposal, 2020).” This rhetoric aligned perfectly with the digital solidarity economy’s goal of using technology to foster democratic ownership and resist corporate data extraction.

Design

From the outset, legal and technical provisions fractured ownership and control, weakening the cooperative. One section of the proposal stated that “data belongs to citizens,” yet another specified that “data obtained with user consent is owned by the city and consortium firms (Internal project proposal, 2020).” Critically, settlement endpoints – such as balance queries and detailed transaction logs – continued to be managed solely by platform administrators, inaccessible to ordinary members. The decision to store cryptographic hashes rather than raw, readable entries on the blockchain made a true community audit impossible. This opaque design played directly into residents’ pre-existing skepticism, shaped by past experiences with non-transparent development projects. As the local project manager P1 recalled, “We saw irregularities before; people read new systems through that lens.”

Practice

On the ground, cooperative organisers bore immense social responsibility without any meaningful control. “We couldn’t ticket illegal parkers; we could only warn them. But people blamed us as if we set the prices or drew the lines,” explained P3. They became the public face of a system governed elsewhere. The information asymmetry fueled corrosive rumors. After a study tour funded by the city, some residents accused an organiser of “pocketing funds,” even though, as she later clarified, “payments went directly from the city to a travel agency; I never touched the money.” The opaque financial and data architecture created the ideal conditions for such mistrust to flourish.

Outcome

Ownership ultimately became symbolic. The cooperative brand existed, local jobs were created, and manuals were written, but the residual rights of control over tariffs, data access, and surplus allocation sat firmly upstream with City P and its vendors. This outcome aligns with Hansmann’s theory. For residents, the governance costs – time spent managing conflicts, the inability to access data, the lack of real power – were exceedingly high and were not offset by tangible benefits. This imbalance made genuine member-ownership untenable and led inevitably to managerial capture by the city and its contractors. The paradox was clear: ownership in narrative, administration in code.

4.3 Solidarity and mutual-aid: Traditions invoked, reciprocity erased

Vision

The project explicitly invoked Korea’s pre-modern mutual-aid traditions of dure (labour pooling), gye (rotating credit), and hyangyak (village compacts) to legitimate its status as a DSSE initiative (Internal project proposal, 2020). By referencing these deeply embedded cultural practices, the city aimed to frame the platform not as a top-down intervention but as a modern expression of organic, historical solidarity. City documents promised 21 local jobs, tokenised reciprocity for community contributions, and a “Mobility Commons Fund” to recycle benefits.

Design

Yet the core instrument of traditional reciprocity – a transparent, member-readable ledger – was entirely absent from the design. The appendices in the project proposal contained detailed staffing plans, operational manuals, and cost-benefit analyses (Internal project proposal, 2020), but no interface where a cooperative member could routinely check a simple statement: “Here is what I contributed, here is what I earned, and here is what the commons fund holds.” The logic of gye and dure depends on this shared visibility to function. The choice to store only hashes reinforced an opacity that was the antithesis of the trust-based systems the project claimed to emulate.

Practice

In practice, solidarity devolved into constant, uncompensated proxy labour for the community’s few active champions. “People have smartphones but don’t really use them. We installed apps and registered IDs for them; even after trainings with big guides and demo videos, they came back and asked us to do it again,” said P3. This technical and emotional burden drained the very community energy the project needed to succeed. Social cohesion also fractured under the strain. “We used to run village markets together; it felt hopeful. After the lines were painted, conflict took over everything,” she recalled.

Outcome

While tangible outputs like jobs and manuals were delivered, the intangible foundation of solidarity – visible reciprocity – was not. P1 reflected that in places scarred by past irregularities, “visibility itself is trust.” He added, “Rumors travel faster than facts.” Without simple, member-level accounting that made mutual benefits visible, rumors and suspicion filled the void. The rhetoric of dure-style solidarity persisted as goodwill among a few, but it failed to crystallise into durable, system-wide cooperation.

4.4 STS and participation performance: Deep tech, thin social interfaces

Vision

The technological backbone promised a suite of “smart” solutions: DID-based identity, Single Sign-On, edge AI for parking enforcement, and a permissioned blockchain for auditable transactions. This vision reflected a powerful state-led sociotechnical imaginary focused on technological solutions as drivers of social progress.

Design

The chosen technology stack, however, systematically deepened the platform’s black box. Specifically, signing up to the system required creating a digital ID and going through several complex authentication steps, which made it hard for less tech-savvy users. Additionally, the blockchain was set up with technical features that only administrators could access, so the promised transparency was, in practice, unavailable to ordinary members.

While technically elegant, these design measures rendered trust invisible and inscrutable to the very community supposed to be in control, perfectly illustrating the infrastructural gap where deep technical layers neutralise formal democratic rights.

Practice

In daily life, residents confronted constant friction, not seamless efficiency. “They came back again and again – we had to do it for them,” noted P3, describing the endless loop of technical support for frustrated elders. Furthermore, the high-tech enforcement system was socially toothless. “We couldn’t issue tickets, only warnings,” an organiser lamented, meaning the AI-powered surveillance had no corresponding social mechanism for enforcement. Residents judged the system not by its aggregate KPIs but by their immediate, frustrating experiences: “Does the app work now? Is parking fair today?”

Outcome

Evaluation reports documented quantifiable improvements – vehicle inflows decreased by 10.6%, illegal parking dropped by more than 100 cases daily, and 21 jobs were generated. However, the final result was a system of promissory transparency and infrastructural solutionism. The back-end was technically capable of auditing transactions, and the key performance indicators (reduced vehicle inflow and illegal parking) were positive, but residents experienced only friction, opacity, and unresolved conflicts. The project’s deep “black-box depth” directly resulted in poor “participation performance,” as the community was alienated rather than empowered by the technology. Without member-readable logs or binding social hooks – like the ability to veto price changes – “trust through code” never became trust in practice.

4.5 Cross-lens synthesis: Where the chain of trust snapped

The project’s slide from an ambitious vision to eventual suspension stemmed from the mutually reinforcing failures that emerged across all four analytical domains described in §2.5. The causal chain began with a powerful sociotechnical imaginary (STS lens) that prioritised a high-tech, efficiency-oriented solution. This imaginary drove the design of a technically complex platform whose deep black-box architecture (e.g., admin-only APIs, hashed data) made genuine community oversight impossible.

This infrastructural opacity directly skewed the calculus of enterprise ownership (Hansmann’s lens). By creating insurmountable information asymmetries, the black box ensured that residents could never become the “lowest-cost controllers,” raising their governance costs to an infinite level. This structure guaranteed managerial capture by the city and its vendors, rendering the cooperative a powerless, symbolic entity.

This failure of ownership, in turn, hollowed out any potential for genuine collaborative governance. With no real control over rules, tariffs, or data, community forums devolved into exercises in performative participation, where residents could voice concerns but not enact change. Finally, this entire structure made the promise of solidarity and mutual-aid untenable. Lacking a transparent ledger to make reciprocity visible, the culturally resonant appeals to dure and gye were exposed as mere rhetoric, unable to survive the daily frictions and conflicts.

The voices of participants confirm this cascading failure. P1’s observation that “we saw irregularities before” highlights the historical mistrust that the black-box design amplified. P11’s demographic point that “elders stay home; interest is low” underscores the mismatch between the high-tech solution and the community’s needs and capabilities. And P3’s lament – “We couldn’t ticket, only warn” – captures the critical disconnect between digital systems and social authority.

The case confirms a crucial lesson for policymakers seeking to build democratic digital infrastructures: binding governance objects, legible public ledgers, and thick social interfaces are not optional add-ons but prerequisites for success. If rules are co-authored before lines are painted, if solidarity is accounted for in member-readable statements, and if deep tech is matched by embedded human support, then the chain of trust might hold. Without them, even the most well-intentioned projects are destined to reproduce the very hierarchies they claim to overcome.

5. Discussion

This section unfolds in three stages. First, in §5.1, we demonstrate how the evidence extends key theories of ownership, collaborative governance, and cooperative platforms. In §5.2, we translate these theoretical insights into concrete design principles for community platforms. Finally, in §5.3, we outline design and policy measures to advance a partner-state approach.

This study set out to investigate the persistent gap between the emancipatory rhetoric of state-sponsored “people-centred” platforms and their on-the-ground institutional and technical realities. By tracing the full life-cycle of the City P shared-mobility project, we found that the platform’s failure was not the result of a single misstep but a cascading series of interconnected failures across governance, ownership, solidarity, and sociotechnical design. The findings offer three major contributions to the literature on platform governance, the Social Solidarity Economy (SSE), and Science and Technology Studies (STS), which this section now unpacks.

5.1 Extending core theories: Ownership, STS, and solidarity

First, this research extends Hansmann’s (1996) enterprise ownership theory to the context of public-sector platform cooperatives. Hansmann’s calculus posits that ownership is most efficient when allocated to the patron group that can govern the enterprise at the lowest cost. In the City P case, a critical new variable emerged: sociotechnical opacity. The platform’s deep “black-box” architecture – from admin-only transaction logs to non-transparent pricing algorithms – drove governance costs so high that cooperative members could never function as Hansmann’s “lowest-cost controllers.” This vacuum was filled by the state and its vendors, confirming the risk of managerial capture by design (Ji, 2018; Sandoval, 2020).

Second, the study operationalises key STS constructs within Korea’s developmental-state context. South Korea’s dominant sociotechnical imaginary portrays science and technology as engines of rapid development, with publics cast as “dutiful citizens” supporting growth-first agendas (Song, 2003; Jasanoff & Kim, 2015). The City P project exemplified this logic. The emphasis on blockchain and AI was not simply technical but political, signalling innovation and progress in line with state 4IR priorities (PCFIR, 2018). Our analysis introduced measures of “black-box depth” and “participation performance” (Paris et al., 2024). We found that depth was technical (e.g., hashed blockchain data) but also legal and procedural (e.g., vague ownership clauses, absence of veto rights). Consequently, participation was hollowed out: residents’ feedback could not alter the platform’s logic, making collaboration largely performative (Ansell & Gash, 2008).

Third, our findings critique the instrumentalisation of cultural heritage in SSE projects. The city invoked dure, gye, and hyangyak to legitimate the platform (Jung & Rösner, 2012; Ji, 2018). Yet these traditions were built on visible reciprocity – labour rotation, transparent ledgers, public compacts (Choi, 2006). A system concealing transactions in an opaque ledger contradicted this core principle. The project thus engaged in a form of infrastructural ventriloquism – using the voice of solidarity without embodying its principles in the system’s functions (Dourish & Bell, 2007). The lesson for SSE is clear: solidarity is not a brand but a practice requiring legible, verifiable reciprocity (Scholz, 2023).

5.2 Cross-lens synthesis, limitations, and future research

Synthesising these points, the project’s trajectory can be seen as a chain reaction. A national imaginary privileged technological sophistication over democratic simplicity (Jasanoff & Kim, 2015), producing a deep black box that blocked oversight (Paris et al., 2024). This opacity prevented genuine ownership, enabling managerial capture (Hansmann, 1996; Ji, 2018). With no real control, collaborative governance collapsed into performative forums (Ansell & Gash, 2008). Without visible reciprocity, solidarity appeals failed (Choi, 2006; Jung & Rösner, 2012).

As a single-case study, these findings are analytically but not statistically generalisable (Yin, 2018). Still, the detailed process-tracing provides comparative leverage. Future work should examine other state-led platforms – both failures and successes – in balancing technological complexity with democratic control.

5.3 Practical design and policy choices for community platforms

From these findings emerge three heuristics for democratic digital commons:

1. Binding governance objects over performative forums: Communities must hold enforceable rights, such as veto powers over tariffs or data-sharing (Ansell & Gash, 2008).
2. Legible public ledgers over promissory transparency: Auditability must be visible to members in everyday interfaces (Choi, 2006; Jung & Rösner, 2012).
3. Thick interfaces over deep tech: Complex systems require human support – help desks, training, non-digital access – to sustain participation (Albornoz, 2023).

A partner-state approach must institutionalise these design commitments. Procurement rules should require open dashboards and resident veto clauses (PCFIR, 2018). Data stewardship should guarantee member-readable ledgers. Capacity-building must fund embedded helpers and accessible channels. In short, democratic principles must be written into contracts, code, and bylaws, or else even cooperative rhetoric will reproduce hierarchy (Sandoval, 2020; Scholz, 2023).

6. Conclusion

This article examined how sociotechnical imaginaries of social value and technological convenience were translated – often mistranslated – into the governance and practice of a state-sponsored platform cooperative. The City P case showed that suspension was not accidental but the predictable outcome of a cascade: a technocratic imaginary favored deep black-box design; opacity displaced community ownership, enabling managerial capture; collaborative forums devolved into performance; and solidarity appeals failed without visible reciprocity (Hansmann, 1996; Jasanoff & Kim, 2015; Song, 2003; Choi, 2006; Jung & Rösner, 2012).

Our contributions are threefold. First, we extend Hansmann’s ownership calculus by demonstrating how sociotechnical opacity functions as a governance cost in the digital era. Second, we operationalise STS concepts – imaginaries, black-box depth, participation performance – in a developmental-state context (Jasanoff & Kim, 2015; Paris et al., 2024). Third, we expose the limits of solidarity-as-branding, showing that traditions like dure and gye require infrastructural encoding to generate trust (Ji, 2018; Choi, 2006).

From these insights follow clear design heuristics and policy steps: binding governance rights, legible ledgers, and thick interfaces, embedded in partner-state arrangements (Ansell & Gash, 2008; PCFIR, 2018; Scholz, 2023). Building truly “people-centred” digital commons demands not only benevolent rhetoric or advanced technology but enforceable transparency, accountability, and reciprocity. Without this alignment, even well-intentioned projects will reproduce the hierarchies they seek to overcome.