Evaluating Independent Bookstore–Public Library–Tourism/Retail Governance: Network Effects and Policy Experiment Design

2.1 Linked governance: structure, management, and effectiveness

Linked governance refers to an interorganizational state in which multiple actors pursue shared purposes under common rules, coordination devices, and accountability routines. In the network-governance literature, three structural ideal types—participant-governed, lead-organization, and network administrative organization (NAO)—capture how authority, coordination capacity, and legitimacy are distributed (Provan & Kenis, 2008). Effectiveness is not presumed; it is contingent on structural fit with network complexity (size, heterogeneity, and interdependence), and on the balance among trust, formal rules, and goal consensus. Public-sector network research further argues that structure and management are coupled with outcomes across multiple levels—community, network, and participating organizations—implying that “success” must be evaluated as a multi-level performance profile rather than a single metric (Provan & Milward, 1995, 2001).

A key implication for bookstore–library–tourism/retail linkages is that governance form is a design choice with trade-offs. Participant governance may strengthen inclusiveness and shared ownership, yet often struggles with coordination costs and decision inertia as the network grows. Lead-organization governance can mobilize resources quickly, but may produce dependency, fatigue, and legitimacy tensions if dominance persists or reciprocity becomes asymmetric (Provan & Kenis, 2008). NAO governance can stabilize routines and evaluation systems, but can also risk over-bureaucratization that dampens grassroots initiative—an issue repeatedly observed in cultural-space policies where institutionalization can inadvertently crowd out bottom-up identities and autonomy (Boswinkel & van Meerkerk, 2023; Magkou, Maud, & Emilie, 2025). Accordingly, linked governance should be theorized through boundary conditions: when and for whom a given structure improves participation, dwell, and revisits, and under what circumstances it generates negative externalities (exclusion, capture, overload, or mission drift).

Recent syntheses further note that “network governance” and “collaborative governance” overlap but are not identical: collaborative governance emphasizes deliberative, cross-boundary co-production under public accountability, while network governance spans broader interorganizational coordination forms including more managerialized arrangements (Wang & Ran, 2023). For this study, the concept of linked governance intentionally bridges these streams by treating governance as a configuration of (a) structural form, (b) relationship-management routines, and (c) performance-management logic. Performance management is particularly salient for evaluation-ready governance design: collaborative arrangements often face measurement ambiguity, indicator gaming, and tensions between learning-oriented and accountability-oriented evaluation regimes (Waardenburg, Groenleer, & de Jong, 2025). Thus, the present framework conceptualizes “evaluation-ready design” as the explicit coupling of governance routines (rules, roles, and calendars) with traceable indicators of participation, dwell, and revisits that can support iterative learning without collapsing the plural values of cultural infrastructure into narrow output metrics.

Finally, process-oriented models of collaboration identify a recurrent pathway from participation conditions to interaction processes, institutional design, and performance, emphasizing regularity, reciprocity, and co-design as operational cores that convert episodic contacts into durable cooperation (Ansell & Gash, 2008; Emerson, Nabatchi, & Balogh, 2012). This study adopts that core but treats it critically: these mechanisms can fail when reciprocity is symbolic, when co-design is nominal, or when regularity becomes ritualized without shared value creation—conditions that can undermine trust and reduce revisit stability despite high program volume.

2.2 Network mechanisms: weak ties, brokerage, and social capital

Network mechanisms specify how linked governance generates outcomes. Weak ties expand reach to new information, audiences, and collaborators and thereby accelerate diffusion beyond existing circles (Granovetter, 1973). Brokerage across structural holes explains how intermediaries connect segmented clusters, create new pathways (“detours”), and recombine resources to produce novel value (Burt, 1992). Social capital denotes relational assets—obligations, trust, norms, and shared expectations—that lower cooperation costs and stabilize coordinated action over time (Coleman, 1988; Putnam, 2000). In bookstore–library–tourism/retail ecosystems, these mechanisms are particularly relevant because the target outcomes—participation, dwell, and revisits—often depend on repeated, low-friction transitions across venues and actors rather than on single-site service quality alone.

However, the same mechanisms can generate unintended consequences if not managed. Weak-tie expansion may broaden exposure but dilute commitment, producing “thin” participation that fails to translate into revisits. Brokerage can accelerate coordination but may also create bottlenecks or gatekeeping if brokerage becomes monopolized by a single actor, increasing vulnerability and dependency. Social capital can stabilize collaboration but may also foster closure that excludes newcomers and reduces innovation. These tensions reinforce why mechanism-based theorizing must incorporate governance design conditions: relationship quality (regularity, reciprocity, co-design) is not merely an intermediate variable but a governing logic that determines whether weak ties and brokerage reduce or amplify pathway resistance from collaboration to measurable outcomes.

Accordingly, this study treats relationship quality as the mechanism-enabling substrate through which weak ties and brokerage become effective. Regularity institutionalizes contact opportunities and makes coordination predictable. Reciprocity sustains symmetric benefits, preventing fatigue and legitimacy erosion. Co-design transforms coordination from exchange to joint production, creating shared narratives and formats that can be repeated, adapted, and scaled. Together, these elements shape whether network structure functions as an “access amplifier” (expanding reach and entry points) and whether it can convert episodic participation into stabilized revisit routes.

2.3 Servicescapes and place: environmental conditions for amplification or attenuation

Servicescape theory argues that behavior, perception, and affect are shaped by configurations of physical, sensory, and social cues (Bitner, 1992). Classic design variables—visibility lines, circulation, density, lighting, seating, acoustics, and micro-zoning—mediate access, dwell, interaction, and the perceived ease of participation. Third-place theory further emphasizes that predictable rhythms and revisit routes are sustained when environments support repeatable rules and comfortable, low-stakes sociability (Oldenburg, 1999). Place theory highlights how regional context, local narratives, and identity-making consolidate experiential meaning (Relph, 1976; Tuan, 1977). Together, these perspectives imply that identical governance devices (e.g., shared calendars, reciprocal passes, reading trails) can yield different magnitudes of effects depending on environmental affordances.

Yet two critical issues require explicit treatment in this study. First, servicescape theory originated in commercial service settings; applying it to public, cultural, and non-commercial infrastructures is contested because “value” is co-produced among heterogeneous stakeholders (users, staff, communities, and local governments) and is often grounded in belonging, care, and civic meaning rather than purchase utility. Recent systematic reviews demonstrate that servicescape scholarship has expanded from physical design cues to multi-dimensional frameworks including social, technological, symbolic, natural, and even spiritual dimensions—particularly relevant for cultural spaces (Fariño Espitia, Rosenbaum, & Contreras-Ramírez, 2025). This study therefore operationalizes servicescape variables (visibility, circulation, dwell-friendliness) as conditional affordances that can amplify or attenuate governance effects, while acknowledging that mechanisms may differ in public cultural infrastructures.

Second, the “infrastructural turn” in cultural policy reframes cultural spaces as infrastructure, but also highlights policy tensions: economic-growth and city-marketing logics can collide with participatory, rights-based, or care-based logics. Research on creative hubs shows that policy often struggles to define what hubs are and risks instrumentalizing them, which can intensify governance frictions and sustainability challenges (Boswinkel & van Meerkerk, 2023). Parallel work on tiers-lieux culturels in France similarly shows controversies around funding priorities, institutional expectations, and the diverse identities of these spaces (Magkou et al., 2025). In library scholarship, public libraries are increasingly theorized as socio-cultural and social infrastructures sustained through everyday “social infrastructuring” practices and continuous care work (van Melik, 2024; Rivano Eckerdal et al., 2024). These debates matter because bookstores and libraries can become policy instruments for tourism or regeneration, but governance designs that ignore place identity and care-based infrastructuring may erode legitimacy and reduce revisit stability. Therefore, the present framework integrates servicescape and place not as context “background,” but as theorized conditions that shape the translation of governance devices into behavioral outcomes.

2.4 Definitions and components

1) Modes of linked governance.

A participant-governed form distributes roles and decision rights across actors under shared rules; a lead-organization form coordinates through a central institution that allocates resources and sets agendas; and an NAO form relies on a dedicated administrative entity to manage network operations and evaluation routines (Provan & Kenis, 2008). In this study, “governance mode” is treated as a structural choice that conditions coordination capacity, legitimacy, and the feasibility of performance management (Provan & Milward, 1995, 2001; Waardenburg et al., 2025).

2) Relationship quality.

Relationship quality is captured by (a) regularity (sustained, scheduled collaboration), (b) reciprocity (balanced exchange and symmetric benefits), and (c) co-design (joint program design and execution). These elements reflect core collaborative-governance process logics and are theorized to translate structural design into operational performance (Ansell & Gash, 2008; Emerson et al., 2012).

3) Network structure.

Network structure is summarized by (a) weak-tie density that widens access to new participants and information, (b) brokerage centrality that links segmented subsets and creates new pathways, and (c) partner diversification that broadens the collaboration spectrum (Granovetter, 1973; Burt, 1992). Structural conditions are not assumed to be uniformly beneficial; they are expected to produce positive effects when coupled with relationship-quality rules that prevent bottlenecks, closure, and fatigue.

4) Environmental affordances

Environmental affordances comprise servicescape factors—visibility, circulation, and dwell-friendliness—that shape user behavior and mediate the translation of governance devices into participation, dwell, and revisits (Bitner, 1992; Fariño Espitia et al., 2025). Place-based meaning and regional context further condition effect size by shaping the salience and resonance of linkage narratives (Relph, 1976; Tuan, 1977).

5) Digital-platform conditions and traceability.

Because linkage governance increasingly operates through digital resources, mobile services, and platform-like coordination devices (shared calendars, membership/pass systems, recommendation channels), digital conditions affect both user adoption and evaluation traceability. Evidence from Korean public-library research shows that impediments to using digital resources—perceived barriers, concerns, and access frictions—can constrain service uptake even when services exist (Noh & Wang, 2021). User-review–based analysis in mobile app markets demonstrates how user-generated data can surface actionable service improvements but also highlights bias and incompleteness when relying on a single platform’s corpus (Noh & Ro, 2021). AI platform policy comparisons further underline that platform strategies and governance choices shape innovation capacity and adoption environments, implying that local cultural-service ecosystems should be designed with attention to platform governance and digital-policy context (Park & Chang, 2022). Accordingly, this study treats digital devices and traceable logs as enabling infrastructure for evaluation-ready design, while recognizing their constraints and potential biases.

2.5 Theoretical propositions

Proposition 1 (Governance introduction and structural fit)

Introducing linked governance among bookstores, libraries, and tourism/retail actors increases participation, dwell, and revisits; the magnitude and stability of effects vary with structural fit relative to network complexity (size, heterogeneity, interdependence) (Provan & Kenis, 2008; Provan & Milward, 1995, 2001).

Proposition 2 (Relationship-quality pathway)

Regularity, reciprocity, and co-design mediate the pathway from governance adoption to outcomes by converting episodic exchanges into repeatable collaboration routines and credible revisit routes (Ansell & Gash, 2008; Emerson et al., 2012).

Proposition 3 (Network-structure amplification with risks)

Higher weak-tie density and brokerage centrality expand information diffusion and access routes, amplifying governance effects when coupled with relationship-quality rules; without such coupling, they may generate bottlenecks, fatigue, or closure that attenuate effects (Granovetter, 1973; Burt, 1992; Coleman, 1988).

Proposition 4 (Servicescape/place moderation)

Given the same governance and mechanism conditions, stronger servicescape affordances—visibility, circulation, and dwell-friendliness—produce larger effects by lowering pathway frictions from participation to dwell and revisits; multi-dimensional servicescape conditions in cultural infrastructures may operate through social and symbolic cues as much as physical design (Bitner, 1992; Fariño Espitia et al., 2025; Oldenburg, 1999).

Proposition 5 (Context heterogeneity and infrastructural politics)

Operative pathways and effect sizes are heterogeneous across regional contexts—tourism intensity, retail structure, public-infrastructure endowments, and digital-platform conditions—such that similar outcomes may emerge via different configurations (equifinality) and may also produce unintended consequences under policy instrumentalization pressures (Relph, 1976; Tuan, 1977; Boswinkel & van Meerkerk, 2023; Magkou et al., 2025; van Melik, 2024; Rivano Eckerdal et al., 2024).

2.6 Domestic scholarship: implications for bookstore–local linkages

Recent Korean studies, often situated in the context of population decline and regional regeneration agendas, highlight bookstores’ emerging public functions and the policy logic of linkage strategies. Domestic scholarship emphasizes bookstores as hybrid cultural spaces, community hubs, and brokers that can connect residents, visitors, and institutions through programs, alliances, and place narratives, thereby providing arguments for recomposing local value through bookstore–library–tourism connections. However, a recurring limitation is that many discussions remain descriptive, with insufficient specification of causal pathways and insufficient differentiation among governance modes and mechanism conditions.

The present paper strengthens this domestic agenda by integrating three additional, directly policy-relevant streams from Korean library and information scholarship. First, evidence on impediments to digital-resource use in Korean public libraries demonstrates that participation growth depends not only on physical linkages but also on reducing digital access frictions and perceived barriers, which can otherwise cap adoption and revisit stability (Noh & Wang, 2021). Second, user-review–based analyses of electronic library services show how user-generated evaluations can be translated into improvement priorities and measurable service redesign targets, while also warning that single-platform corpora may introduce bias—an insight directly aligned with this paper’s emphasis on evaluation-ready design and transparent indicatorization (Noh & Ro, 2021). Third, comparative analyses of AI platform policies and strategies highlight that platform governance and innovation-capacity conditions shape adoption environments and policy feasibility, implying that local cultural ecosystems should be designed with sensitivity to platform-like coordination, data governance, and policy context (Park & Chang, 2022).

International cultural policy debates reinforce why this integration matters. Creative hubs and tiers-lieux research shows that policy support can be simultaneously enabling and destabilizing if it imposes narrow growth logics, redefines space identities, or crowds out bottom-up practices (Boswinkel & van Meerkerk, 2023; Magkou et al., 2025). Library scholarship on socio-cultural infrastructure and social infrastructuring further argues that sustainable outcomes depend on continuous care practices and everyday work that do not always register in conventional output indicators (van Melik, 2024; Rivano Eckerdal et al., 2024). Finally, research on libraries as cultural and educational tourism hubs suggests that tourism linkage potential is real but demands careful evaluation design to avoid over-claiming effects and to manage trade-offs between visitor attraction and local community value (Rejeki et al., 2024). Together, these literatures imply that the central gap is not the lack of “linkage ideas,” but the lack of a transferable, evaluable framework that specifies how governance modes, relationship-quality rules, network structures, environmental affordances, and digital-platform conditions jointly produce participation, dwell, and revisits—while also anticipating failure modes and unintended consequences.

The present study addresses this gap by proposing an evaluation-ready conceptual framework that couples (i) governance mode, (ii) relationship quality, (iii) network structure, and (iv) environmental affordances (including digital traceability) as an integrated design language for bookstore–library–tourism/retail ecosystems.

3.1 Design overview

This study is purpose-built to evaluate the effects of governance linking independent bookstores, public libraries, and tourism/retail using only a published fieldwork corpus. All analyses rely exclusively on the report’s narrative text, tables, photo captions, maps, and appendices (Noh & Lee, 2024a), following procedures that are reproducible without new fieldwork, surveys, or administrative data. The study has two aims: first, to systematically operationalize text-derived, evaluation-ready indicators that proxy participation, dwell, and revisits; and second, to identify pathways and conditions by which linked governance leads to participation, dwell, and revisits by combining within-case process tracing and cross-case configurational comparison via fuzzy-set Qualitative Comparative Analysis (fsQCA) (Bennett & Checkel, 2015; Ragin, 2008; Schneider & Wagemann, 2012). The emphasis is on identifying rules, pathways, and conditions grounded in the corpus, rather than estimating absolute effect sizes.

3.2 Corpus and unit of analysis

The corpus is restricted to the complete report on independent bookstores in the Gangwon and Chungcheong regions (Noh & Lee, 2024), and the primary unit of analysis is the individual bookstore case as described in the report. Within each case, episodes—adoption of library pick-up linkages, the start of joint programs, participation in inter-bookstore campaigns, pilots of reciprocal memberships/passes—are reconstructed in narrative context using temporal markers such as “thereafter,” “regularly,” and “settled.” For regional contrasts, Gangwon and Chungcheong serve as auxiliary strata; however, indicators and rules are derived from textual evidence at the bookstore-case level.

3.3 Operationalization of concepts

Governance modes are classified into participant-governed, lead-organization, and network administrative organization (NAO) forms (Provan & Kenis, 2008). In-text cues such as “operating council,” “standing coordination meeting,” and “overall coordination/secretariat” guide assignment. Relationship quality is defined along three dimensions: regularity (sustained scheduled collaboration), reciprocity (balanced exchange and mutual benefit), and co-design (joint design and execution of programs). These are coded from cue combinations such as “weekly/monthly/institutionalized,” “reciprocal/ cross-benefit,” and “co-designed/co-hosted” (Ansell & Gash, 2008; Emerson, Nabatchi, & Balogh, 2012).

Network structure is captured by the density of weak ties that widen access to new participants and information, brokerage centrality that links segmented subsets, and partner diversification as the spectrum of collaborating actors. Institution–activity co-mentions are extracted to build a two-mode matrix and projected to a one-mode network for indicator calculation (Granovetter, 1973; Borgatti & Everett, 1997; Burt, 1992).

Environmental affordances comprise servicescape factors—visibility, circulation, and dwell-friendliness—derived from descriptions of seating, lighting, acoustics, stage placement, sightlines, and modular combinations (café, gallery, workshop, archive). These are ordered on an ordinal intensity scale (Bitner, 1992).

Outcome variables are participation, dwell, and revisits. Expressions such as “fully booked, standing room only, serial courses,” “staying/studying/working/seat use,” and “regulars/clubs/standing meetings,” combined with regularity markers, are converted into low–medium–high ordinal indicators. All indicators are designed as text-minimal proxies that permit comparison across cases using the corpus alone.

Because all measures are derived from a single documentary corpus, the indicators should be interpreted as comparative proxies rather than direct, absolute estimates of effect size. To mitigate narrative and selection biases, we (i) retain an audit trail of coding decisions and exception rules, (ii) register refutational or countervailing passages as negative evidence, and (iii) triangulate across narrative body, captions, maps, and appendices to reduce single-source over-reliance. We also frame causal claims conservatively: the analysis identifies plausible mechanisms and configurations consistent with the propositions, while acknowledg:contentReference[oaicite:21]{index=21}tion and unobserved confounding.

3.4 Data extraction and coding procedure

Source materials include the report’s narrative body, captions, maps, and appendices. A domain dictionary—e.g., “regular/weekly/monthly/serial/season,” “joint/collaborative/ alliance/linked,” “library/school/festival/tourism,” “seating/lighting/acoustics/circulation/stage/sightline,” “regulars/club/meeting”—is used for batch retrieval, followed by contextual verification. Captions and map annotations, which are high-salience cues for relationships and space, are prioritized. Co-mentioned institution–activity pairs are concurrently logged to form the raw two-mode network table.

Reliability is strengthened by piloting to refine the codebook, optional double coding with Krippendorff’s α where feasible, and, under a single-coder design, scheduled recoding to assess stability (Krippendorff, 2013; Miles, Huberman, & Saldaña, 2014). Coder training notes, decision rules, and exception handling are fully documented as an audit trail (Lincoln & Guba, 1985).

3.5 Network construction and indicator calculation

Based on co-mentions between institutions (bookstores, libraries, schools, festivals, local governments, associations) and activities (programs, campaigns, calendars, memberships), a two-mode matrix is constructed and projected onto a one-mode institution–institution network (Borgatti & Everett, 1997). Edge weights are assigned by the number of joint participations or mention frequency. At the node level, degree and betweenness centrality are computed; at the network level, modularity and patterns of segmentation/cohesion are examined (Freeman, 1979; Opsahl, Agneessens, & Skvoretz, 2010). Increases in brokerage centrality are interpreted as filling structural holes and creating detour routes (Burt, 1992), while higher weak-tie density indicates expanded ingress for new participants and information (Granovetter, 1973). These indicators are aligned with the propositions in Chapter 2 to assess mediating and moderating structures that connect governance design to participation, dwell, and revisits.

3.6 Identification strategy and comparative design

Within cases, temporal markers and event episodes are stitched to reconstruct sequences of introduction–settling–scaling, tracing changes in expressions related to participation, dwell, and revisits before and after adoption of linkage devices (Bennett & Checkel, 2015). Across cases, linkage devices, relationship quality, network structure, and environmental affordances are calibrated as fuzzy sets to probe sufficient and necessary configurations; equifinality is assessed via consistency and coverage (Ragin, 2008; Schneider & Wagemann, 2012). Because all measures are derived from a single documentary corpus, the indicators should be interpreted as comparative proxies rather than direct, absolute estimates of effect size. To mitigate narrative and selection biases, we (i) retain an audit trail of coding decisions and exception rules, (ii) register refutational or countervailing passages as negative evidence, and (iii) triangulate across narrative body, captions, maps, and appendices to reduce single-source over-reliance. We also frame causal claims conservatively: the analysis identifies plausible mechanisms and configurations consistent with the propositions, while acknowledg:contentReference[oaicite:21]{index=21}tion and unobserved confounding. Where clear introduction points exist—e.g., the launch of an alliance campaign—pre–post shifts in expression frequency are additionally presented as an event-style contrast, while explicitly acknowledging the limits of within-document pre–post comparisons (Abadie, Diamond, & Hainmueller, 2010).

3.7 Robustness, ethics, and transparency

Robustness checks proceed in three steps. First, cut-points for ordinal-to-fuzzy calibration are varied within reasonable ranges to test rule stability. Second, edge-weighting schemes are alternated across mention counts, event counts, and event-type diversity to examine sensitivity. Third, cases with multiple devices introduced in the same period are labeled “joint introduction” and interpreted separately during process tracing.

On ethics, the study analyzes publicly available publications and contains no personally identifying information. For transparency and reproducibility, the codebook, domain dictionary, calibration rules, network-construction scripts, and a registry of refutational cases are archived as an audit trail (Lincoln & Guba, 1985; Miles et al., 2014).

4.1 Coding results

First-cycle in vivo coding followed by second-cycle categorization across the entire corpus converged on expression clusters directly related to linked governance: regularity, reciprocity, co-design, the expansion of weak ties, the emergence of brokerage roles, and fine-grained descriptions of servicescape affordances. Temporal markers of regularity and cadence—“weekly,” “monthly,” “institutionalized,” “serial”—and joint attributions—“co-hosted,” “collaborative,” “alliance”—recurred in many program descriptions. Connections to libraries, schools, festivals, and local creators tended to be narrated not as isolated events but as time-series flows. Photo captions and map annotations offered rich cues on the coupling of spatial modules and on dwell-friendly features such as sightlines, circulation, and seating; these descriptions were often linked to the intensity of event formats and implied the accumulation and acceleration of experience (Bitner, 1992).

4.2 Trends in identifying governance modes

Cues such as “operating council,” “standing coordination meeting,” and “overall coordination/secretariat” signaled the presence of network administration functions, and in some periods cases exhibited structures approaching lead-organization or NAO forms. By contrast, the operation of many day-to-day programs bore the hallmarks of participant governance, with schedules, spaces, and resource allocation realized through partner agreement and role sharing rather than heavy concentration in a single institution (Provan & Kenis, 2008). When structures temporarily shifted from participant-governed to lead-organization, the narrative typically coincided with alliance campaigns or region-wide events, followed by a reversion to participant governance during routine operations (Noh & Lee, 2024).

4.3 Relationship quality: regularity, reciprocity, and co-design

Expressions such as “regular meetings,” “seasonal operation,” and “serial courses/projects” indicated regularity, while terms like “mutual benefit,” “cross-benefit,” and “reciprocal pick-up” functioned as indicators of reciprocity. Devices including library pick-up for requested titles, joint promotion and shared calendars, and cross-venue use simultaneously embodied the design and execution of symmetric relations. Traces of co-design appeared in co-hosting labels, role sharing in program design, and the use of alliance branding; where these elements combined, descriptions of participation and revisits tended to be positive (Ansell & Gash, 2008; Emerson, Nabatchi, & Balogh, 2012; Noh & Lee, 2024). In higher-quality relationships, narratives emphasized not one-off cooperation but the embedding of ties, aligning with arguments on network sustainability (Provan & Milward, 2001).

4.4 Network structure: weak ties and the rise of brokers

Based on co-mentions of institution–activity linkages in the corpus, connections to libraries, schools, festivals, and local creators suggested an expansion of weak ties that broadened ingress for new participants and information (Granovetter, 1973; Noh & Lee, 2024). During intervals marked by inter-bookstore initiatives or regional campaigns, specific bookstores or institutions were portrayed as intermediaries stitching together program linkages, consistent with brokerage that fills structural holes (Burt, 1992). Where brokerage roles were explicit, narratives also tended to note partner diversification and a widening program spectrum, alongside frequent references to new visitors or external inflows.

4.5 Environmental affordances: visibility, circulation, and dwell-friendliness

Assured sightlines during events, the relative placement of stage and seating, separated or connected circulation paths, calibrated lighting and acoustics, and modular combinations of café, gallery, workshop, and archive were described as compound devices that promote dwell (Noh & Lee, 2024). In servicescape terms (Bitner, 1992), these elements lower behavioral friction for access, dwell, and interaction and operate as mediators and moderators of program effects. Cases rich in dwell-friendly descriptions showed smoother transitions across formats—reading groups, workshops, exhibitions/talks—where spatial convenience undergirded the intensity and diversity of programming.

4.6 Typologies of operative pathways

Comparative reading of the corpus yielded three representative pathways. First, the library-linked pathway broadened low-barrier participation via library pick-up, joint promotion, and co-programming, then transitioned to regular meetings and serial courses that stabilized revisits. Second, the alliance/campaign pathway raised visibility through shared branding and circulating calendars among multiple bookstores and partners, then fanned out into each bookstore’s distinctive formats for deepening. Third, the local-integration pathway strengthened motives for dwell by connecting with local authorship, craft, festivals, and tourist itineraries, composing hybrid experiences. These pathways were not mutually exclusive; narratives depicted overlap and shifts over time, with pathway transitions commonly mediated by organized brokerage capacity and the inertia of co-design (Noh & Lee, 2024).

4.7 Regional variation and conditional effects

In Gangwon, alliance programming and ties to tourism itineraries appeared relatively prominent, with frequent references to external inflows and seasonal linkages. In Chungcheong, regular learning communities centered on children and creation were emphasized, together with role transitions among insiders and the institutionalization of serial projects. The same devices tended to yield effects via partner diversification and amplified external visibility in Gangwon, and via accumulated regularity and co-design in Chungcheong (Relph, 1976; Tuan, 1977; Noh & Lee, 2024). These patterns suggest that the magnitude and pathways of governance effects are moderated by the place-specific regional context.

4.8 Counterexamples and boundary conditions

Where program menus were broad but narratives of regularity or role rules were thin, references to revisits and community consolidation were comparatively scarce. Conversely, even with limited program diversity, repeated linkages with libraries and schools or returns to alliance campaigns sustained mentions of participation inflows and heightened visibility. This indicates that outcomes are more sensitive to the combination of relationship quality, network structure, and environmental affordances than to program variety per se (Bitner, 1992; Provan & Milward, 2001). In addition, when lead-organization forms appeared to be fixed over longer periods, some passages hinted at partner dependence and fatigue, underscoring the need to adapt governance modes to situational fit (Provan & Kenis, 2008; Noh & Lee, 2024).

4.9 Summary

The corpus-based findings support rules whereby the adoption and quality of linked governance, together with network structure and environmental affordances, mediate and moderate participation, dwell, and revisits. Repeatedly, experiences accumulated and accelerated, and outcomes stabilized when low-barrier formats expanded through co-design and partner diversification and when regular rhythms converged with spatial affordances. Variation between Gangwon and Chungcheong demonstrates multiple pathways to similar results, implying the necessity of localizing linked-governance design. Linking results to propositions. Proposition 1 is supported by repeated narrative shifts in participation, dwell, and revisits following the introduction of linkage devices, with stronger effects observed where governance structure fit the scale and diversity of partners. Proposition 2 is evidenced by the co-occurrence of regularity, reciprocity, and co-design cues with stabilized revisits and embedded ties, indicating that relationship quality functions as a key pathway rule. Proposition 3 is supported by increased partner diversification and explicit intermediary roles during alliance/campaign periods, consistent with weak-tie expansion:contentReference[oaicite:25]{index=25}. Proposition 4 is reflected in cases where visibility, circulation, and dwell-friendly spatial modules lowered “pathway frictions,” enabling smoother transitions from low-barrier participation to deepening formats. Proposition 5 is corroborated by regional heterogeneity: similar outcomes emerged via different pathways across contexts, underscoring the need for localized design rather than one-size-fits-all governance prescriptions.

5.1 Integrative interpretation

The corpus-based analysis indicates that the pathways through which linked governance increases participation, dwell, and revisits are jointly mediated and moderated by relationship quality, network structure, and servicescape affordances (Granovetter, 1973; Burt, 1992; Bitner, 1992; Provan & Milward, 2001). In particular, regularity, reciprocity, and co-design function as rules that convert one-off exchanges into embedded relations, while weak ties and brokerage broaden external reach and facilitate inflows of new participants and information (Granovetter, 1973; Burt, 1992). Because environmental attributes such as visibility, circulation, and dwell-friendliness alter pathway frictions and thus amplify or dampen outcomes, the conjunction of spatial/social environments with network design emerges as pivotal (Bitner, 1992). Differences between Gangwon and Chungcheong—an “alliance/tourism” trajectory versus a “learning-community rhythm” trajectory—show that distinct routes can converge on similar results (high experiential and participatory density), a pattern consistent with equifinality (Noh & Lee, 2024).

5.2 Principles for governance design

Three design principles follow. First, codify schedules, spaces, and roles into a regular rhythm. Institutionalization supplies predictability and accumulates learning and relationships with minimal complexity (Oldenburg, 1999; Provan & Kenis, 2008). Second, formalize symmetric reciprocity. Devices such as library pick-up for requested titles, joint promotion and a shared calendar, and cross-venue use preserve balance of benefits and support durability of cooperation (Provan & Milward, 1995, 2001). Third, adopt co-design as the default mode. Role division at the planning stage and shared branding strengthen participant identification and enable the reuse of repeatable formats (Emerson, Nabatchi, & Balogh, 2012; Ansell & Gash, 2008). Fourth, specify and staff brokerage. During alliance periods or regional campaigns, heightened betweenness accompanies partner diversification and program expansion; institutionalizing the actor or unit that fills structural holes is therefore warranted (Burt, 1992).

5.3 Joint optimization of servicescapes and programs

The servicescape is both an amplifier and a stabilizer of program effects (Bitner, 1992). Clear sightlines, the relative placement of stage–seating–stacks, minimized circulation conflicts, calibrated lighting and acoustics, and modular combinations (café, gallery, workshop, archive) smooth transitions from low-barrier to deepening formats. When spatial affordances couple with regular rhythms, narratives repeatedly report longer dwell and stabilized revisit routes. Governance design should thus presume dual optimization: rules for schedules and roles in tandem with adjustments to visibility, circulation, and dwell pockets.

5.4 Implementation roadmap and a minimum viable bundle

A practical minimum bundle comprises a shared calendar, reciprocal membership/pass, library pick-up linkage for requested titles, and a reading trail. The calendar supplies visibility and predictability; the reciprocal pass offers symmetric benefits and revisit incentives; pick-up linkages provide a standing touchpoint between libraries and bookstores; and reading trails connect offerings to local movement patterns. Bundling yields stronger effects than piecemeal adoption. During alliance campaigns, a temporary shift to a lead-organization or NAO form is appropriate; routine operations can revert to participant governance (Provan & Kenis, 2008). The broker oversees calendar operations, partner matching, and joint promotion, and periodic feedback meetings correct drifts from reciprocity and co-design.

5.5 Building evaluation into operations

Comparability using text alone was feasible because narratives consistently reported regularity, joint attributions, partner diversification, and spatial affordances. In practice, assigning unique identifiers and standardized caption rules to programs/events/co-designed initiatives—and recording co-host labels and partner lists consistently—would allow pathway- and condition-level comparisons using minimal text-based indicators. Lightweight participation logs embedded in the shared calendar and membership system can further link relationship quality (regularity, reciprocity, co-design) and network structure (brokerage centrality, diversification) to outcomes with greater transparency (Provan & Milward, 2001). Such “evaluation-ready design” also eases subsequent mixed-methods integration with administrative, transaction, or mobility data.

5.6 Context-contingent strategies

In tourism-dense regions, shared branding, circulating calendars, and reading trails yield strong visibility effects, and the key task is converting one-time external visits into stable revisit routes. Where learning-community rhythms are prominent, the cumulative force of regular meetings and the inertia of co-design are decisive; reciprocal passes and pick-up linkages strengthen continuous touchpoints between libraries and bookstores (Noh & Lee, 2024; Relph, 1976; Tuan, 1977). In both settings, institutionalized brokerage and a sustained balance of reciprocity are prerequisites for performance.

5.7 Risks and boundary conditions

A prolonged, fixed lead-organization form can generate partner fatigue and dependence, weakening diversification (Provan & Kenis, 2008). When co-design remains nominal, real role division and reciprocity erode, undermining trust (Provan & Milward, 1995). Expanding program menus without improving spatial affordances may disperse dwell and interaction, failing to stabilize revisits (Bitner, 1992). Finally, text-based comparisons are vulnerable to narrative bias and omission; standardized recording rules and simple activity logs during operations are advisable to bolster interpretive stability.

5.8 Policy recommendations

Actionable implications for research design and measurement. Future studies can make the evaluation agenda more actionable by combining (i) a core set of comparable indicators (participation, dwell, revisit, partner diversification, brokerage capacity, and rhythm regularity) with (ii) context-sensitive indicators (tourism seasonality, local retail ecology, and public infrastructure endowments). Indicator validity can be strengthened through “lightweight” triangulation—e.g., aligning proxy scores with program logs, reservation/pick-up records, footfall counters, or mobility traces—while retaining the low-cost comparability of the present :contentReference[oaicite:27]{index=27}ring negative cases (breakdowns, asymmetry, fatigue, dependence) as systematically as success cases would sharpen boundary conditions and improve policy transferability.

Stakeholder engagement and adaptive governance. For implementation, a minimal governance protocol can specify (a) who holds the broker role, (b) how reciprocity is audited (symmetric benefit checks), (c) how often joint calendars and rosters are updated, and (d) how disputes and overload are handled. Periodic “design reviews” with libraries, bookstores, tourism actors, and local government can institutionalize learning and prevent drift toward either excessive centralization or fragmentation. This aligns the paper’s design rules with collaborative performance management concerns, improving both accountability and adaptability.

Statements and Declarations

Author Contributions: Conceptualisation, Y.N.; methodology, Y.N.; software, Y.N.; validation, Y.N.; formal analysis, Y.N..; investigation, Y.N.; resources, Y.N.; data curation, Y.N.; writing—original draft preparation, Y.N.; writing—review and editing, Y.N.; visualisation, Y.N.; supervision, Y.N.; project administration, Y.N.; funding acquisition, Y.N. All authors have read and agreed to the published version of the manuscript.

Funding: This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea(NRF-2023S1A5C2A02095114)

Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of Konkuk University (IRB: KKUIRB-202503-HR-050).

Informed Consent Statement: Informed consent was obtained from all participants involved in the study.

Data Availability Statement: The data presented in this study are available on reasonable request from the corresponding author.

Acknowledgments: We would like to thank Editage (www.editage.co.kr) for English language editing. We would also like to thank the industrial safety experts who assisted with this research.

Conflicts of Interest: The authors declare no conflict of interest.