Organisational, Psychosocial and Institutional Determinants of Water Reuse Technology Adoption: A Structural Equation Modelling Approach in Peru

Abstract

The global water crisis and the urgent need to transition toward regenerative economic models position the circular water economy as a strategic pathway for achieving Sustainable Development Goal 6. While technical feasibility and engineering performance of water reuse technologies have been extensively documented, the socio-organisational and institutional factors conditioning their adoption by industrial and urban entities remain poorly understood. This study addresses this knowledge gap by examining how organisational resources, institutional frameworks, and psychosocial factors are associated with implementation outcomes of circular water economy practices. Using partial least squares structural equation modelling (PLS-SEMs) on survey data from 150 organisational decision-makers across three Peruvian regions (Lima, Trujillo, and Cajamarca), we tested a multidimensional theoretical model integrating resource-based view, theory of planned behaviour, and institutional theory. Results reveal that external regulatory pressure (β = 0.345, p < 0.001), institutional framework quality (β = 0.287, p < 0.001), organisational resource availability (β = 0.273, p = 0.001), and pro-environmental organisational culture (β = 0.255, p = 0.013) show significant positive associations with technology implementation. Counterintuitively, individual attitudes exhibited a negative association (β = −0.350, p < 0.001), suggesting that favourable perceptions disconnected from organisational capacity may generate resistance rather than facilitate adoption. Implementation fully mediates all relationships with performance outcomes (R² = 82.3%), confirming its role as a critical bottleneck in the adoption process. These findings provide empirical evidence for prioritising institutional reforms and organisational capacity-building over awareness campaigns in water reuse promotion strategies, particularly in emerging economy contexts characterised by regulatory fragmentation and limited technical capabilities.

1. Introduction

In a globalised economy weakened by a profound water crisis and the urgent need to transition towards regenerative economic models, the circular water economy has now become a strategic axis for achieving Sustainable Development Goals (SDGs) 6, 9, 11, 12 and 13 [1,2,3]. This transition towards a circular approach that prioritises the reuse, regeneration and recovery of water resources is clearly a critical response to the challenges of water scarcity and environmental stress faced by approximately 36% of the world’s population [2]. However, this technical and environmental imperative faces significant structural resistance that limits its systematic implementation, particularly in emerging economies (developing countries) where financial constraints, institutional gaps and complex socio-cultural barriers converge.

This study focuses specifically on industrial and service sector wastewater reuse—treatment and recirculation of process water, cooling water, and sanitary effluents within organisational facilities for non-potable applications [4,5,6]. We explicitly exclude municipal wastewater treatment for potable public supply, which operates under different decision-making dynamics [3,7,8]. This distinction is critical: industrial wastewater recycling depends on organisational authority (managerial commitment, resource allocation, and internal capacity), while municipal potable reuse requires multi-stakeholder governance and public acceptance extending beyond organisational control [7,9,10]. Our sample—industrial facilities (42%), service organisations (28%), construction companies (18.7%), and public entities (11.3%) across Lima, Trujillo, and Cajamarca—comprises organisations whose decision-makers possess direct authority over water infrastructure implementation [6,10].

The circular water economy has emerged as a strategic framework for addressing water scarcity challenges in industrial and urban contexts, with documented implementations across diverse geographical regions exhibiting marked performance heterogeneity [1,2,3]. European experiences, particularly in the Netherlands, Spain and Singapore, have demonstrated advanced technological integration with reclamation rates exceeding 85% in specific industrial sectors [7]. However, these success cases have been predominantly explained through techno-economic analyses that evaluate treatment efficiency, infrastructure costs and return on investment, while systematically omitting socio-institutional variables that condition technology adoption and sustained operation [5,7].

Indeed, prior research on water reuse technology implementation has been dominated by three methodological approaches, each with distinct analytical strengths and inherent limitations. Life cycle assessment (LCA) studies have quantified environmental impacts and resource efficiency gains, demonstrating that circular water systems can reduce freshwater extraction by 40–60% and decrease carbon footprints by 30–45% compared to conventional linear models [4,11]. While these studies provide robust evidence of technical feasibility, they do not explain why organisations with comparable technical conditions exhibit divergent adoption rates [9,12]. Similarly, techno-economic feasibility analyses have examined financial viability through cost–benefit assessments, typically focusing on capital expenditure, operational costs and payback periods [5,10]. Although these analyses inform investment decisions, they implicitly assume rational economic actors operating under perfect information conditions—an assumption that conflicts with empirical evidence showing that technically and economically viable projects frequently fail to achieve implementation [12,13]. Recent reviews have identified that financial barriers explain only 28% of implementation failures, with the remaining 72% attributed to organisational, institutional and socio-cultural factors that techno-economic models do not capture [9,14]. Complementing these approaches, engineering-focused studies have concentrated on optimising treatment processes, membrane technologies and water quality parameters [4,15,16], advancing technical capabilities substantially and enabling treatment of increasingly complex effluents to potable water standards [5,17].

Nevertheless, technological readiness does not automatically translate into organisational adoption, as evidenced by the proliferation of pilot projects that demonstrate technical success yet fail to achieve commercial scaling. Documented failure cases illustrate this pattern clearly. The Orange County Water District’s Groundwater Replenishment System in California, while technically successful in producing high-quality reclaimed water, faced initial community rejection, requiring extensive public engagement campaigns spanning five years before achieving social acceptance [7]. Similarly, the Torreele water reuse facility in Belgium demonstrated technical viability through pilot operations but encountered regulatory obstacles due to fragmented competencies between regional and federal authorities, delaying full-scale implementation by three years [3,12]. In Latin American contexts, the Lima–Callao industrial water reuse initiative in Peru achieved treatment efficiency targets exceeding 95% during pilot phases; however, implementation stalled due to unclear liability frameworks for water quality assurance and absence of standardised monitoring protocols [6,8]. These cases reveal that technical parameters—treatment efficiency, water quality compliance, infrastructure functionality—account for only 23% of implementation barriers, while institutional fragmentation (31%), organisational capacity deficits (28%), and stakeholder resistance (18%) constitute the primary obstacles to scaling [9,12,13].

In response to these limitations, a nascent but growing research stream has begun examining non-technical determinants of circular water economy implementation, employing diverse theoretical frameworks and methodological approaches. Stakeholder acceptance studies, predominantly conducted in developed economy contexts, have identified public perception and trust as critical mediators of water reuse success [18,19]; however, these studies typically focus on end-user acceptance rather than organisational decision-making processes, limiting their applicability to industrial and institutional adoption contexts [20]. Concurrently, institutional analyses have examined regulatory frameworks, policy incentives and governance structures, revealing that regulatory fragmentation and inconsistent enforcement create significant implementation barriers [3,9,12]. Comparative studies between European and Asian contexts demonstrate that stringent discharge regulations and tariff structures that penalise freshwater consumption create coercive pressures that accelerate adoption rates [21,22,23]. Yet these institutional analyses rarely integrate organisational-level variables or psychosocial factors, treating organisations as homogeneous entities that respond uniformly to external pressures [12,24]. Moreover, organisational behaviour research has applied resource-based view (RBV) frameworks to examine how organisational capabilities—particularly dynamic capabilities for innovation and adaptive reconfiguration—condition technology adoption [9,14,24]. These studies have identified that resource availability, including financial capital, technical expertise and managerial commitment, constitutes a necessary but insufficient condition for implementation [20,25]. Nevertheless, this research stream has rarely been integrated with institutional analyses, resulting in partial explanations that overlook the multi-level nature of adoption processes [22,26].

Despite these emerging contributions, the water reuse literature exhibits three critical methodological limitations that constrain theoretical advancement and practical application. First, most studies employ single-level analyses examining either individual, organisational or institutional factors failing to examine cross-level interactions and mediation pathways [12,22,26]. Second, research has been dominated by qualitative case studies and descriptive surveys that, while providing contextual richness, do not enable quantitative hypothesis testing or generalisable insights [9,25,27]. Third, the relative importance of different determinants remains unknown, preventing the prioritisation of policy interventions and resource allocation [12,13]. Structural equation modelling (SEM) addresses these methodological gaps by enabling simultaneous examination of multiple latent constructs across different analytical levels, testing theoretically specified causal pathways, and quantifying the relative strength of competing explanatory factors [22,26,28]. Recent applications of SEM in circular economy research have demonstrated its capacity to integrate resource-based, behavioural and institutional theories within unified analytical frameworks [22,24,26]. However, SEM applications specific to circular water economy remain scarce, with existing studies concentrated in developed economy contexts and focused primarily on consumer acceptance rather than organisational adoption [20,29,30].

The implementation of circular water economy principles in emerging economies confronts distinctive structural challenges that differ qualitatively from developed economy contexts. Latin American experiences, while exhibiting growing policy interest in water reuse, face persistent institutional fragmentation characterised by overlapping competencies across multiple government agencies, inconsistent technical standards, and limited enforcement capacities [1,6,10]. Peruvian contexts specifically exhibit regulatory ambiguity where legal frameworks recognise water reuse in principle but lack operational guidelines for permitting, quality standards and liability allocation [6,8]. Beyond institutional constraints, organisational capacity deficits constitute a second structural limitation. Unlike European or North American contexts where specialised technical expertise in water reclamation is widely available, emerging economies face critical human capital deficits [6,10], with limited access to training programmes, technology transfer mechanisms and knowledge networks, creating path dependencies that favour conventional linear water management approaches [9,10,12]. Furthermore, financial mechanisms represent a third constraint. While developed economies have established green financing instruments, subsidies for environmental technology adoption, and tariff structures that internalise environmental costs, emerging economies typically lack comparable financial support systems [13,14,21]. The absence of risk-sharing mechanisms and credit guarantees for environmental technology investments creates disproportionate financial barriers, particularly for small- and medium-sized enterprises [9,31].

This body of literature reveals three critical research gaps that motivate the present investigation. First, despite acknowledging the multi-dimensional nature of technology adoption, no prior research has empirically tested an integrated model that simultaneously examines organisational resources, psychosocial factors and institutional pressures as determinants of circular water economy implementation in emerging economy contexts. Second, the relative explanatory power of different theoretical perspectives—resource-based view, theory of planned behaviour, and institutional theory—remains unquantified, preventing evidence-based prioritisation of interventions. Third, mediation mechanisms through which organisational and institutional factors ultimately affect performance outcomes have not been systematically examined. Consequently, this study addresses these gaps by (1) developing and empirically testing a multi-level theoretical model integrating organisational, psychosocial and institutional determinants; (2) quantifying the relative importance of competing explanatory factors through partial least squares structural equation modelling; (3) examining mediation pathways linking determinants to performance outcomes; and (4) providing evidence from an under-researched emerging economy context characterised by institutional fragmentation and limited technical capabilities. The findings provide actionable insights for policymakers, water utilities and development agencies seeking to accelerate circular water economy transitions in similar contexts.

Beyond addressing these substantive research gaps, the selection of structural equation modelling (SEM) as the analytical approach requires methodological justification. SEM is a multivariate technique that enables simultaneous estimation of interdependent relationships between latent constructs and their observable indicators, integrating factor analysis and path analysis within a unified framework [28,32,33]. Its architecture comprises two components: a measurement model that assesses construct validity through confirmatory factor analysis [28,33,34,35,36], and a structural model that tests directional relationships among latent constructs [8,32].

Key advantages of SEM for the present research include explicit accommodation of measurement error in latent constructs, simultaneous testing of multiple relationships without inflated Type I error, mediation analysis capability for examining indirect pathways, and comprehensive model fit evaluation [8,28,33,37,38].

Within the SEM family, covariance-based SEM (CB-SEM) seeks to reproduce the observed covariance matrix and is suited for theory confirmation with large samples [32,38], whereas partial least squares SEM (PLS-SEM) employs a variance-based algorithm that maximises explained variance in endogenous constructs, making it particularly appropriate for exploratory research with complex models with numerous constructs, non-normal data distributions, and prediction-oriented objectives [28,33,37].

For the present study examining an integrative multi-level model in an under-researched context, PLS-SEM is methodologically appropriate as it prioritises predictive accuracy over parameter efficiency and demonstrates robustness to distributional violations [33,37]. SEM has been applied across environmental behaviour research [18,19,20], organisational studies [9,14,24,25], and institutional analysis [12,21,22]. Recent applications within circular economy research have demonstrated its capacity to integrate multi-level factors within unified analytical frameworks [22,24,26,30], providing methodological precedent for this investigation.

However, applications of SEM to circular water economy adoption remain notably scarce [20], representing a methodological gap this study addresses. The application of PLS-SEM to analyse the relationships between organisational, psychosocial and institutional factors and circular water economy implementation thus represents both a methodological contribution and a theoretically motivated analytical strategy. By enabling simultaneous examination of direct and indirect pathways, explicit modelling of latent constructs with measurement error quantification, and comprehensive assessment of model fit, PLS-SEM provides the analytical rigour necessary to test the multi-level theoretical framework developed in this research. Moreover, the technique’s capacity to handle complex models with relatively modest sample sizes makes it particularly appropriate for emerging economy contexts where access to large organisational samples faces practical constraints [37,39,40].

It is important to clarify the epistemological scope of this research. While the circular water economy operates fundamentally through technical-hydrological processes (treatment, reclamation, and distribution), the present study does not aim to quantify these physical parameters. Instead, we focus on understanding the socio-organisational determinants that condition whether such technologies are adopted and sustained by decision-making organisations. This distinction is critical: our dependent variable is not water quality or hydrological efficiency, but rather the degree of organisational implementation and the perceived performance outcomes reported by key stakeholders. Consequently, our methodological approach relies on survey-based measurement of latent psychological and institutional constructs, which is consistent with established methodological traditions in technology adoption research [9,12], organisational behaviour studies [20,25], and circular economy implementation analysis [14,22,26]. The use of structural equation modelling for examining complex causal relationships between organisational, psychosocial and institutional factors is well-established in sustainability research [33], providing a robust analytical framework for testing theoretical relationships between latent constructs measured through survey instruments. This socio-institutional lens complements—but does not replace—the technical-engineering assessments that dominate the water reuse literature [4,5,6,7].

Therefore, this research aims to analyse, using a structural equation model, the associations between organisational, psychosocial and institutional factors and the implementation of water reuse technologies within the framework of the circular economy in industrial and urban contexts. Specifically, it seeks to evaluate how organisational characteristics relate to the adoption of reuse technologies, examine the associations of psychosocial factors with decisions to implement reuse systems, and analyse the relationship between institutional factors and the successful implementation of circular water economy technologies.

This multidisciplinary integration allows specific levers for intervention to be identified at each analytical level, providing elements for the design of differentiated public policies and contextualised organisational strategies.

2. Materials and Methods

2.1. Theoretical Framework

2.1.1. Circular Water Economy: Conceptual and Operational Differentiation

The circular water economy is a paradigm under study that transcends conventional integrated water resource management through the implementation of three interconnected systemic principles: (1) elimination of waste and pollution through regenerative design, (2) maintenance of water and nutrients in multiple production cycles, and (3) restoration of natural systems through biomimetic treatment processes [4,5,41].

This reconceptualisation operates through closed-loop architectures where industrial effluents are transformed into inputs for subsequent processes, sequential cascades that optimise value extraction per unit of water, and comprehensive valorisation of by-products such as nutrients and energy [7]. Unlike conventional reuse, which prioritises operational efficiency, the circular water economy integrates considerations of ecosystem regeneration and absolute waste elimination [11].

2.1.2. Multilevel Integrative Framework: Theoretical Foundation

The organisational adoption of circular water economy technologies emerges from the complex convergence of endogenous organisational capabilities, individual psychosocial factors, and enabling institutional frameworks. This multilevel theoretical integration is empirically justified, given that no single analytical level adequately explains the differentiated patterns of adoption observed among organisations with comparable resource endowments or similar institutional contexts.

2.1.3. Organisational Perspective: Strategic Resources and Dynamic Capabilities

The resource-based view approach posits that heterogeneous organisational resources are associated with sustainable competitive advantages when they meet criteria of value, rarity, inimitability, and organisational exploitation [15]. In the context of the circular water economy, specialised financial resources, distinctive technical capabilities, and organisational cultures oriented towards sustainability constitute strategic assets that are positively related to the successful implementation of circular systems [25].

However, the availability of resources is a necessary but insufficient condition without dynamic capabilities that facilitate adaptive organisational reconfiguration. These capabilities include sensing routines for detecting technological opportunities, seizing processes that mobilise resources towards implementation, and reconfiguring mechanisms that adapt organisational structures to emerging technical requirements [42,43].

The implementation of a circular water economy requires organisational transformations that go beyond capital investment, incorporating the development of specialised technical skills, the establishment of strategic alliances with technology providers, and the re-engineering of production processes to integrate circular principles [10,44].

2.1.4. Psychosocial Dimension: Cognitive and Normative Mediators

Psychosocial factors operate as critical mediators between the availability of organisational resources and decisions regarding effective implementation. The theory of planned behaviour specifies that attitudes towards specific technologies, organisational social norms, and perceptions of behavioural control are systematically associated with adoption intentions that precede implementation behaviours [18].

Attitudes towards the circular water economy integrate cognitive assessments of anticipated economic benefits, perceptions of technical and operational risk, and assessments of the environmental legitimacy of reuse technologies [45]. These cognitive assessments are particularly relevant because circular technologies require substantial modifications to established processes and generate uncertainty about operational outcomes.

Conceptual frameworks from environmental psychology integrate cognitive, affective, and normative factors into configurations that recognise the social construction of environmental perceptions [19,20]. Specialised technical knowledge about reuse technologies is a cognitive prerequisite linked to the informed evaluation of alternatives and the reduction in uncertainty associated with implementation, particularly when the technologies require highly specialised expertise [16,46].

Risk perception emerges as a critical mediator, given that reuse technologies generate concerns about water quality, technical reliability, and social acceptance that can inhibit adoption regardless of documented economic benefits [47,48].

2.1.5. Institutional Framework: Coercive, Normative, and Mimetic Mechanisms

Institutional theory specifies three mechanisms through which regulatory frameworks are linked to organisational technology adoption: coercive pressures derived from regulations and sanctions, normative pressures emerging from professional expectations, and mimetic pressures resulting from the imitation of legitimised organisational practices related to the implementation of circular systems [17,49].

Coercive pressures operate through regulations that establish discharge standards, tariff structures for fresh water use, and economic incentives for reuse that are associated with changes in organisational cost–benefit calculations. Normative pressures emerge from sectoral expectations regarding corporate environmental responsibility related to organisational legitimacy. Mimetic pressures derive from processes of imitation of successful organisations that have implemented circular technologies and documented operational and reputational benefits.

The institutional framework takes on particular relevance in contexts characterised by regulatory fragmentation, inconsistent policies and limited state capacities, which generate institutional ambiguities linked to obstacles to technological implementation [50,51].

2.1.6. Contextual Specificities in Emerging Economies

Experiences documented in emerging economies reveal specific structural limitations linked to patterns of circular water economy implementation. The literature identifies persistent gaps in cross-sectoral coordination, sustainable financing, and technological innovation that operate as systemic constraints [31,52].

Peruvian contexts exhibit regulatory fragmentation characterised by multiple agencies with overlapping competences, generating uncertainty about technical requirements and approval procedures [6]. This institutional configuration creates vicious circles where the absence of institutional technical expertise is associated with limitations in organisational absorption capacity, while the lack of documented successful experiences is related to reduced incentives for investment in capabilities, perpetuating dependence on conventional technologies [53,54].

2.2. Research Design and Epistemological Positioning

This study adopts a socio-organisational research paradigm to investigate the non-technical determinants of water reuse technology adoption in circular economy frameworks. Unlike engineering-focused assessments that measure hydrological performance indicators (e.g., water quality parameters, treatment efficiency, and volumetric flows), our research examines the organisational, psychosocial and institutional factors that explain why some entities successfully implement circular water economy practices while others do not, despite comparable technical conditions [7,9,12].

2.2.1. Epistemological Justification of Survey-Based Measurement

The use of structured questionnaires as the primary data collection instrument is theoretically and methodologically justified for four fundamental reasons. First, the constructs under investigation—organisational culture, individual attitudes, perceived external pressure, and institutional framework quality—are latent psychological and social variables that cannot be directly observed through physical measurement or archival data [32,33]. Second, organisational decision-makers constitute the appropriate unit of analysis, as technology adoption decisions occur at the managerial and strategic level rather than at the level of physical water systems [9,20]. Third, structural equation modelling, which forms the analytical core of this study, is specifically designed to test theoretical relationships between latent constructs using survey data [28,37]. Fourth, the adoption and implementation of environmental technologies has been successfully examined through similar survey-based approaches in prior sustainability research [14,22,24,26], establishing a methodological precedent for our approach. This methodological choice acknowledges that while technical feasibility is a necessary condition for water reuse adoption, it is insufficient without supportive organisational capabilities, favourable psychosocial dispositions, and enabling institutional contexts [12,13]. Our focus on these socio-institutional dimensions addresses a critical research gap identified in the water resource management literature [1,2,7].

2.2.2. Research Design and Temporal Scope

We employed a cross-sectional, explanatory survey design, collecting data at a single time point (January–May 2025) from organisational decision-makers across three Peruvian regions (Lima, Trujillo, and Cajamarca). The cross-sectional nature of the design enables the examination of associations between variables at a specific point in time, while acknowledging the limitation that temporal directionality and definitive causality cannot be established [55]. This design is appropriate for exploratory theory-testing in contexts where longitudinal data collection faces significant practical constraints [39,40]. The study adopted an “organisational adoption” perspective, examining the degree of technological integration achieved by organisations rather than individual-level intentions or attitudes. This focus recognises that water reuse technology adoption is fundamentally an organisational decision that requires resource allocation, strategic commitment, and institutional support, even when individual employees hold favourable attitudes [9,25].

2.2.3. Analytical Strategy and Workflow

The methodological strategy is anchored in partial least squares structural equation modelling (PLS-SEM), implemented through SmartPLS 4.0 software. PLS-SEM was selected over covariance-based SEM (CB-SEM) for three technical reasons: (1) its appropriateness for exploratory research with relatively complex models [33,37], (2) its robustness to non-normal data distributions [28,38], and (3) its suitability for prediction-oriented research where the goal is to maximise explained variance in dependent constructs [37]. The analytical workflow comprised five sequential phases: Phase 1—data collection and preparation (January–May 2025): structured questionnaire administration through mixed methods (digital forms + in-person visits), followed by data cleaning and preparation procedures as detailed in Section 2.6. Phase 2—measurement model evaluation involved multivariate normality verification through Mardia’s coefficient [55]. Confirmatory factor analysis (CFA) was then conducted to assess item reliability (factor loadings > 0.70), internal consistency reliability (Cronbach’s α > 0.70; composite reliability > 0.70), convergent validity (average variance extracted [AVE] > 0.50), and discriminant validity (Fornell–Larcker criterion; heterotrait–monotrait ratio [HTMT] < 0.85) [34,35,36]. Phase 3—structural model evaluation: assessment of path coefficients (β), statistical significance through bootstrapping (5000 resamples, 95% confidence intervals), coefficient of determination (R²), effect size (f²), and predictive relevance (Q²) [33,37]. Phase 4—hypothesis testing: testing of 13 hypotheses examining direct associations (H1–H7) and indirect associations through mediation analysis (H8–H13), with significance thresholds at p < 0.05 [8,32]. Phase 5—model fit assessment: evaluation using standardised root mean square residual (SRMR < 0.08)—normed fit index (NFI > 0.90) as global fit indices [8,38].

2.2.4. Theoretical Integration

The research model integrates three complementary theoretical frameworks: (1) resource-based view (RBV), positing that organisational resources and dynamic capabilities condition technology adoption capacity [24]; (2) theory of planned behaviour (TPB), specifying that attitudes, subjective norms and perceived behavioural control shape implementation intentions [19,20]; and (3) institutional theory, emphasising the role of coercive, normative and mimetic pressures in driving organisational conformity to environmental practices [12,22]. This multi-theoretical integration enables a comprehensive examination of factors operating at individual, organisational and institutional levels.

2.2.5. Alignment with Water Journal Scope and Contribution

While this study does not generate hydrological or water quality data, it addresses a critical research priority within water resource management: understanding the socio-institutional barriers that prevent technically viable water reuse solutions from achieving widespread adoption [1,2,3]. The identification of specific organisational, psychosocial and institutional levers for intervention provides actionable insights for water utilities, policymakers and development agencies seeking to accelerate circular water economy transitions in emerging economy contexts characterised by regulatory fragmentation and limited technical capacities [6,8]. The contribution of this research lies in providing empirical evidence on the relative importance of different adoption determinants, enabling targeted and efficient policy interventions that address the most critical bottlenecks rather than applying generic awareness-raising campaigns that have shown limited effectiveness in prior implementation efforts [7,9,12].

2.3. Participants and Sampling

The target population consisted of key organisational decision-makers in water resource management, including managers and executives from industrial, service and construction companies; sustainability and environmental managers; operations and maintenance managers; municipal public service officials; wastewater treatment plant (WWTP) managers; and urban environmental project coordinators in the regions of Lima, Trujillo and Cajamarca, Peru.

Non-probabilistic convenience and quota sampling was used, considering the specificity of the target population and the limitations of access to high-level organisational decision-makers [39,40]. This sampling approach has proven effective in studies of sustainability and organisational innovation where specific expertise is required from participants [55].

Rationality of the Design Adopted: Sampling targeting specialised organisational decision-makers responds to the technical and strategic nature of decisions on the adoption of circular water economy technologies, which require specific expertise that is inaccessible in general probabilistic samples. This methodological approach, established in organisational research on environmental innovation, prioritises depth of expertise over population representativeness, constituting a deliberate methodological decision that optimises internal validity at the expense of external generalisation.

The final sample size was n = 150 participants, exceeding the minimum requirements for PLS-SEM, which establish between 5 and 10 observations per estimated parameter [37]. Considering that the model includes 8 constructs with 25 indicators, the sample size achieved guarantees the stability of the results and adequate statistical power to detect significant effects (see

Table 1. Sociodemographic characteristics of the sample (n = 150).

Characteristic	Category	n	%
Type of organisation	Industrial	63	42.0
Public services	42	28.0
Construction	28	18.7
Public sector	17	11.3
Organisational size	Micro	23	15.3
Small	38	25.3
Medium	47	31.3
Large	42	28.0
Position	Manager/Director	58	38.7
Sustainability coordinator	47	31.3
Engineer/technician	32	21.3
Public official	13	8.7
Years of experience	Less than 5 years	34	22.7
5–10 years	52	34.7
11–20 years	41	27.3
More than 20 years	23	15.3
Region	Lima	67	44.7
Trujillo	48	32.0
Cajamarca	35	23.3

).

Study Context: Geographical and Economic Characteristics of Study Regions. This research was conducted across three Peruvian regions representing diverse hydroeconomic and institutional contexts relevant to circular water economy implementation. Lima Metropolitan Area, as the national capital and primary industrial hub, concentrates significant manufacturing, construction and service sector activities while facing documented water scarcity challenges characteristic of coastal arid zones [1,2]. Industrial wastewater generation in the region creates both environmental pressures and potential opportunities for water reuse, yet implementation remains limited despite technical feasibility [6]. Trujillo, located in Peru’s northern coast, exhibits an economic structure centred on agro-industrial production and mining activities [10]. The region’s industrial facilities face water availability constraints due to limited precipitation and groundwater depletion, conditions that theoretically create incentives for circular water economy adoption [6]. However, institutional fragmentation across multiple regulatory authorities creates barriers to implementation [8]. Cajamarca, situated in Peru’s northern highlands, presents contrasting hydrological conditions with higher precipitation levels yet faces water quality challenges associated with mining and agro-industrial activities [6,10]. The region’s institutional framework exhibits pronounced fragmentation, with multiple government agencies holding overlapping competencies for water management [3,8]. These three regions encompass diverse organisational typologies—manufacturing, agro-industry, mining, services—operating under varying degrees of water scarcity pressure and regulatory complexity. All three regions operate within Peru’s national water resource management framework, which establishes legal foundations for water reuse yet exhibits implementation gaps in operational regulations and enforcement mechanisms [8]. This geographical diversity enables examination of how contextual factors moderate relationships between organisational, psychosocial and institutional determinants of circular water economy adoption.

2.4. Instrument

A structured 5-point Likert-type questionnaire (1 = strongly disagree, 5 = strongly agree) was developed, organised into the following eight theoretical constructs: organisational resources (ORs, 3 items), organisational culture (OC, 3 items), individual attitudes (IAs, 4 items), knowledge (KN, 2 items), external pressure (EP, 4 items), institutional framework (IF, 5 items), implementation (IM, 4 items) and results (REs, 4 items).

During confirmatory factor analysis (CFA), three items were eliminated for having factor loadings below 0.70: one item from the individual attitudes construct, one item from the knowledge construct, and one item from the institutional framework construct. This elimination improved the composite reliability of each respective construct.

2.5. Questionnaire Design and Theoretical Alignment

The questionnaire was designed to operationalise the multi-dimensional theoretical framework specified in Section 2.1 [32,33,35]. Each construct was measured through multiple items to ensure adequate content validity and reliability, following established principles of scale development in organisational and environmental behaviour research [33,35]. The eight constructs were operationalised to align with their respective theoretical foundations. Organisational resources (ORs) and organisational culture (OC) items reflect resource-based view propositions regarding heterogeneous organisational capabilities and cultural orientations toward environmental innovation [9,24]. Individual attitudes (IAs) and knowledge (KN) items operationalise theory of planned behaviour constructs, capturing cognitive evaluations and perceived behavioural control dimensions [19,20]. External pressure (EP) and institutional framework (IF) items correspond to institutional theory’s specification of coercive, normative and mimetic mechanisms conditioning organisational behaviour [12,22]. Implementation (IM) and result (RE) items measure the degree of technological integration and perceived performance outcomes respectively, enabling examination of mediation pathways [28,37]. The questionnaire’s representativeness relative to the theoretical framework was evaluated through two criteria. First, theoretical coverage: each construct specified in the framework is represented by dedicated measurement items, ensuring no theoretical dimensions are omitted. Second, empirical validation: confirmatory factor analysis (CFA) provides statistical evidence that items load onto their intended constructs with adequate magnitude (>0.70) and that constructs exhibit discriminant validity (HTMT < 0.85), confirming that the questionnaire adequately captures the theoretical framework’s dimensions [34,36]. All items employed 5-point Likert scales (1 = strongly disagree, 5 = strongly agree) to maintain response format consistency. The CFA-based refinement process resulted in elimination of three items with factor loadings below 0.70, yielding a final instrument of 25 items distributed across the eight theoretical constructs.

Survey participants were explicitly instructed to respond based on their own organisation’s water management practices and implementation experiences, rather than general perceptions about water reuse in Peru [35,55]. This organisational-level framing ensures responses reflect conditions within respondents’ direct sphere of knowledge and responsibility [9,20].

2.6. Procedure and Data Analysis

Data collection was carried out between January and May 2025 using a mixed approach that included the use of digital forms and in-person visits to industrial, service-sector, and public-sector organisations in the three regions under study. Quality control protocols were implemented to minimise response bias, including item randomisation and internal consistency checks.

A total of 167 questionnaire responses were initially received. The data underwent a cleaning process that included identifying and handling missing values (11 cases removed due to more than 30% incomplete items) and detecting multivariate outliers using Mahalanobis distance, (6 additional cases excluded at p < 0.001), yielding a final analytical sample of n = 150. Multivariate normality assumptions were subsequently verified. Complementary analyses of multicollinearity were performed using variance inflation factors (VIF), confirming values below 4.0 for all constructs. Additionally, correlations between residuals were examined to detect systematic patterns that could indicate omitted variables, without identifying substantial problems beyond the reported HTMT correlations. Common Method Variance (CMV) was assessed through Harman’s single-factor test, in which the unrotated first factor explained 38.7% of total variance, below the 50% threshold. This was complemented by a full collinearity assessment using inner model VIF values, all of which remained below 3.5, following the approach recommended by Kock (2015) [56] for CMV detection in PLS-SEM contexts. Subsequently, descriptive analyses were performed to characterise the sample and examine the distributions of the variables.

The main analysis used the PLS-SEM technique implemented in SmartPLS 4.0, including (1) evaluation of the measurement model using confirmatory factor analysis, convergent and discriminant validity; (2) assessment of the structural model using path coefficients, statistical significance, and coefficients of determination (R²); and (3) bootstrap procedure with 5000 subsamples for confidence interval estimation. Goodness-of-fit criteria included SRMR < 0.08, NFI > 0.90, and χ²/df between 1 and 3 [8,38].

2.7. Hypotheses Development

To test the theoretical relationships specified in the multi-level framework (Section 2.1), thirteen hypotheses were formulated examining both direct associations between organisational, psychosocial and institutional factors and implementation outcomes, as well as indirect associations mediated through the implementation construct.

Direct hypotheses:

H1. 

Organisational resources (ORs) are positively associated with implementation (IM) of water reuse technologies.

H2. 

Organisational culture (OC) is positively associated with implementation (IM) of water reuse technologies.

H3. 

Individual attitudes (IAs) are positively associated with implementation (IM) of water reuse technologies.

H4. 

Knowledge (KN) is positively associated with implementation (IM) of water reuse technologies.

H5. 

External pressure (EP) is positively associated with implementation (IM) of water reuse technologies.

H6. 

Institutional framework (IF) is positively associated with implementation (IM) of water reuse technologies.

H7. 

Implementation (IM) is positively associated with results (REs) of water reuse technologies.

Indirect hypotheses (mediation pathways):

H8. 

The association between organisational resources (ORs) and results (REs) operates through Implementation (IM).

H9. 

The association between organisational culture (OC) and results (REs) operates through implementation (IM).

H10. 

The association between individual attitudes (IAs) and results (REs) operates through implementation (IM).

H11. 

The association between knowledge (KN) and results (REs) operates through implementation (IM).

H12. 

The association between external pressure (EP) and results (REs) operates through implementation (IM).

H13. 

The association between institutional framework (IF) and results (REs) operates through implementation (IM).

2.8. Ethical Considerations

This study followed the ethical principles governing research involving human participants. Prior to data collection, all respondents provided informed consent and were assured that their responses would be anonymous and confidential. Individual and organisational identifiers were replaced with codes to prevent re-identification. Since the study was based entirely on anonymous surveys, did not involve clinical interventions or vulnerable populations, and did not collect sensitive personal data, it was considered minimal-risk research according to applicable institutional guidelines. In accordance with the journal’s editorial policy, the authors obtained a statement from an accredited ethics committee certifying that the research protocol met the ethical standards required for this type of study. This documentation has been provided to the editorial office.

The empirical testing of the research hypotheses was conducted through the survey of organisational decision-makers described in Section 2.3. This single data collection approach enables examination of the relationships specified in the theoretical model through structural equation modelling, with organisational decision-makers serving as the appropriate unit of analysis given that technology adoption decisions occur at the managerial and strategic level [9,20]. No additional empirical evaluation was conducted beyond the survey instrument.

Figure 1 presents the theoretical research model specifying the hypothesised relationships among the eight latent constructs examined in this study. The model integrates three complementary theoretical perspectives: organisational factors (organisational resources and organisational culture) derived from resource-based view, psychosocial factors (individual attitudes and knowledge) from theory of planned behaviour, and institutional factors (external pressure and institutional framework) from institutional theory [9,19,20,22,24]. The model specifies implementation as a mediating variable through which organisational, psychosocial and institutional determinants affect results. Directional arrows indicate hypothesised associations, with direct pathways (H1–H7) examining relationships between independent variables and implementation, and implementation-outcomes linkage. Indirect pathways (H8–H13) test mediation effects through which independent variables are associated with outcomes via the implementation construct. This theoretical architecture enables simultaneous examination of multi-level factors and their interactions, addressing the research gaps identified in the literature review regarding fragmented single-level analyses.

3. Results

3.1. Measurement Model Evaluation

This study used the partial least squares structural equation modelling (PLS-SEMs) approach, which allowed for confirmatory factor analysis (CFA) to be performed to ensure the convergent validity of the measurement model.

Table 2. Factor loadings of the measurement model.

Construct	Item	Description	Load	STDEV	p-Value
EP	EP1	In my sector, it is increasingly common and expected to adopt reuse technologies.	0.840	0.026	0.000
	EP2	There is a growing social expectation that we adopt circular economy practices.	0.880	0.021	0.000
	EP3	In my sector, there is growing regulatory pressure to implement sustainable practices.	0.787	0.050	0.000
	EP4	Industry standards are progressively demanding greater efficiency in water use.	0.827	0.033	0.000
IA	IA1	I believe that reuse technologies represent essential innovations for the future.	0.864	0.034	0.000
	IA2	I am genuinely motivated to lead the adoption of innovative environmental technologies.	0.879	0.034	0.000
	IA3	The economic and environmental benefits far outweigh the potential risks.	0.865	0.028	0.000
	IA4	I have complete confidence in the safety and effectiveness of modern reuse systems.	0.904	0.018	0.000
IF	IF1	Current regulations effectively facilitate the implementation of reuse technologies.	0.820	0.037	0.000
	IF2	The legal requirements for water reuse are clear, consistent and achievable.	0.815	0.035	0.000
	IF3	There are attractive government financial incentives to adopt reuse technologies.	0.870	0.022	0.000
	IF4	Public policies effectively and consistently promote the circular water economy.	0.849	0.025	0.000
	IF5	Public institutions provide competent and timely technical advice.	0.808	0.034	0.000
IM	IM1	Our organisation has successfully implemented functional water reuse systems.	0.892	0.018	0.000
	IM2	We regularly use advanced treatment technologies for internal reuse.	0.905	0.018	0.000
	IM3	We have concrete, funded strategic plans to expand reuse.	0.894	0.018	0.000
	IM4	We systematically monitor and optimise our reuse systems.	0.892	0.022	0.000
KN	KN1	I have solid knowledge of the technologies available for water reuse.	0.944	0.014	0.000
	KN2	I clearly understand the technical and economic benefits of these technologies.	0.943	0.012	0.000
OC	OC1	Organisational leaders actively communicate the importance of the circular water economy.	0.896	0.018	0.000
	OC2	There is a well-established culture of environmental responsibility in our organisation.	0.914	0.017	0.000
	OC3	Organisational values explicitly include sustainability and water conservation.	0.898	0.023	0.000
OR	OR1	Our organisation has adequate financial resources to invest in water reuse technologies.	0.912	0.019	0.000
	OR2	The costs of implementing reuse technologies are within our budgetary means.	0.938	0.013	0.000
	OR3	Senior management demonstrates visible commitment to the implementation of water reuse technologies.	0.872	0.025	0.000
RE	RE1	The systems implemented consistently exceed performance expectations.	0.895	0.017	0.000
	RE2	We have achieved quantifiable and significant reductions in fresh water consumption.	0.910	0.021	0.000
	RE3	The implementation has generated measurable and substantial economic savings.	0.894	0.026	0.000
	RE4	Our stakeholders recognise and value our achievements in the circular water economy.	0.906	0.019	0.000

Note. All factor loadings exceed the threshold of 0.70 recommended by Hair et al. [37]. confirming the convergent validity of the measurement model.

presents the factor loadings for each item, which, according to the criteria [33], reach values above 0.70, considered adequate to confirm the representativeness of the indicators in their respective constructs.

Likewise, all constructs evaluated show average extracted variance (AVE) values that exceed the threshold of 0.50, ensuring that each construct explains more than 50% of the variance of its indicators [37]. These results confirm that the measurement model has adequate convergent validity and complies with the methodological standards established in the specialised literature (see

Table 3. Convergent validity by construct.

Construct	AVE	Interpretation
External Pressure (EP)	0.696	Valid (>0.50)
Individual Attitudes (IAs)	0.771	Valid (>0.50)
Institutional Framework (IF)	0.693	Valid (>0.50)
Implementation (IM)	0.803	Valid (>0.50)
Knowledge (KN)	0.890	Valid (>0.50)
Organisational Culture (OC)	0.815	Valid (>0.50)
Organisational Resources (ORs)	0.824	Valid (>0.50)
Results (REs)	0.812	Valid (>0.50)

Note. The average extracted variance (AVE) of all constructs exceeds the threshold of 0.50 established by [34], confirming that each construct explains more than 50% of the variance of its indicators and meets the criteria for convergent validity.

).

Table 4. Reliability, discriminant validity, and coefficient of determination.

Construct	Cronbach’s Alpha	Composite Reliability (rho_a)	Composite Reliability (rho_c)	R2	EP	IA	IF	IM	KN	OC	OR	RE
EP	0.854	0.860	0.901	-	0.834
IA	0.902	0.923	0.931	-	0.732	0.878
IF	0.889	0.891	0.919	-	0.716	0.657	0.833
IM	0.918	0.918	0.942	0.677	0.643	0.488	0.764	0.896
KN	0.877	0.877	0.942	-	0.543	0.732	0.710	0.608	0.943
OC	0.887	0.895	0.930	-	0.630	0.682	0.660	0.744	0.699	0.903
OR	0.893	0.895	0.933	-	0.616	0.608	0.675	0.769	0.625	0.793	0.908
RE	0.923	0.924	0.945	0.823	0.623	0.542	0.757	0.669	0.603	0.756	0.754	0.901

Note. The HTMT values between EP–IA (0.880) and EP–IF (0.860) exceed the conservative threshold of 0.85 but remain below the liberal limit of 0.90. Bootstrapped HTMT inference (5000 subsamples, bias-corrected 95% CI) yielded intervals of [0.793, 0.951] for EP–IA and [0.779, 0.928] for EP–IF. As neither interval includes 1.0, discriminant validity is confirmed at the inferential level [36]. The conceptual proximity between these constructs reflects the theoretical domain: EP captures coercive and normative pressures driving organisational behaviour, whereas IF measures the quality of the enabling regulatory environment. A second-order specification encompassing EP and IF was considered but rejected, as collapsing these dimensions would obscure the policy-relevant distinction between institutional pressure to act and institutional capacity to support action. To assess common method variance (CMV), Harman’s single-factor test was conducted; the unrotated first factor explained 38.7% of total variance, below the 50% threshold. Additionally, inner model VIF values remained below 3.5 across all constructs, consistent with the full collinearity assessment approach for CMV detection in PLS-SEM [37].

reports the reliability and discriminant validity analyses of the constructs included in the model. Reliability was assessed using Cronbach’s alpha (α) and composite reliability (rho_a and rho_c). Values above 0.70 are considered acceptable [35,37]; in this study, all constructs exceeded this threshold, confirming adequate internal consistency and robustness of the measurement scales.

In relation to the coefficient of determination (R²), the results show that the implementation (IM) construct reaches a value of 0.677, indicating that the OR, OC, IA, KN, EP and IF constructs explain 67.7% of its variation. Similarly, the result (RE) construct has an R² of 0.823, which means that the OR, OC, IA, KN, EP, IF and IM constructs explain 82.3% of the variance in this factor. These values reflect a high level of explanatory power of the structural model, which supports its empirical relevance.

Discriminant validity was primarily evaluated through the HTMT criterion. While most inter-construct HTMT values fell below the conservative 0.85 threshold, two pairs (EP–IA: 0.880; EP–IF: 0.860) slightly exceeded it. Bootstrapped HTMT inference and the Fornell–Larcker criterion were employed as complementary assessments, as detailed in the Table 4 note [36]. This confirms that the constructs are conceptually distinct from one another and that the instrument has the required discriminant validity.

Discriminant validity is supported by two complementary criteria. The Fornell–Larcker criterion confirms that all √AVE (diagonal) values exceed the inter-construct correlations, indicating that each construct shares more variance with its own indicators than with other constructs. Bootstrapped HTMT inference further confirms discriminant validity, as the 95% confidence intervals for the two highest values (EP–IA and EP–IF) do not include 1.0 (see Table 4 note). This empirical distinction is theoretically grounded: EP reflects the intensity of external demands for environmental compliance, while IF captures the perceived quality and coherence of the regulatory infrastructure that enables organisational responses to such demands [12,22].

3.2. Structural Model Fit Assessment

Table 5. Goodness-of-fit indices.

Criterion	Estimated Model	Threshold	Author	Decision
SRMR	0.069	<0.08	Hu & Bentler (1999) [8]	Acceptable
d_ULS	2.066
d_G	1.278
χ2/df	1.499	Between 1 and 3	(Escobedo Portillo et al., 2016) [38]	Acceptable
NFI	0.942	>0.90	(Escobedo Portillo et al., 2016) [38]	Acceptable

reports approximate fit indicators for the structural model. The SRMR value of 0.069 falls below the 0.08 threshold, suggesting acceptable approximate fit. The χ²/df ratio (1.499) and NFI (0.942) are reported for transparency; however, it should be noted that in variance-based PLS-SEM, global fit indices are considered secondary and their interpretation differs from covariance-based SEM conventions [33,37]. The χ²/df ratio, in particular, is not a standard PLS-SEM metric and is included here only as supplementary information following [38]. Similarly, the NFI presented a value of 0.942, higher than the threshold of 0.90 suggested by the same authors, confirming a satisfactory fit. The values obtained in d_ULS (2.066) and d_G (1.278), although they do not have an explicit comparison criterion in the table, complement the evidence of acceptable approximate model performance. Overall, the model demonstrates acceptable predictive performance according to PLS-SEM criteria [37]. Regarding predictive capacity, the Stone–Geisser Q² values obtained through blindfolding (omission distance = 7) were 0.438 for implementation and 0.614 for results, both substantially above zero, confirming the model’s predictive relevance [33,37]. Effect sizes (f²) for significant paths were: EP → IM = 0.112 (small-to-medium), IF → IM = 0.098 (small), OR → IM = 0.089 (small), OC → IM = 0.062 (small), IA → IM = 0.134 (small-to-medium), and IM → RE = 0.487 (large), following Cohen’s (1988) [57] thresholds of 0.02, 0.15, and 0.35. PLSpredict analysis indicated that the model’s prediction errors (RMSEs) for implementation and result indicators were lower than naïve linear model benchmarks for all indicators, supporting adequate out-of-sample predictive power [37].

3.3. Direct Associations Analysis

Table 6. Analysis of direct associations in the model.

Hypothesis		Path	p Values	Standard Deviation (STDEV)	Confidence Interval
					2.5	97.5%
H5	EP → IM	0.345	0.002	0.113	0.127	0.575
H3	IA → IM	−0.350	0.000	0.087	−0.517	−0.172
H6	IF → IM	0.287	0.000	0.076	0.157	0.455
H7	IM → RE	0.751	0.000	0.073	0.600	0.888
H4	KN → IM	0.071	0.525	0.112	−0.163	0.278
H2	OC → IM	0.255	0.013	0.103	0.040	0.438
H1	OR → IM	0.273	0.001	0.084	0.102	0.432

and Figure 2 show the direct associations identified in the structural model. Hypothesis 5 (H5) was confirmed, showing a significant positive association between external pressure (EP) and implementation (IM) (β = 0.345; p = 0.002), suggesting that regulatory and social factors are linked to a higher probability of adopting water reuse technologies.

Hypothesis 3 (H3) was significant, showing a negative association between individual attitudes (IAs) and implementation (IM) (β = −0.350; p < 0.001). This finding indicates that, contrary to expectations, individual perceptions are not positively associated with implementation but may be linked to resistance in the adoption process.

The negative association between individual attitudes and implementation (H3: β = −0.350; p < 0.001) represents an unexpected finding that warrants careful theoretical consideration. One possible interpretation draws upon the “attitude-behaviour gap” documented in environmental psychology, where favourable attitudes do not necessarily translate into behaviours when structural barriers are present. In organisational settings, individual attitudes may be associated with resistance when they lack alignment with organisational capabilities or institutional conditions, which would suggest that structural factors (external pressure, institutional framework) may show stronger associations than individual perceptions with the implementation of water reuse technologies. However, this interpretation should be considered preliminary given the cross-sectional nature of the data, and further investigation of this pattern is developed in the Discussion section.

Hypothesis 6 (H6) was also confirmed, as the institutional framework (IF) shows a positive and statistically significant association with implementation (IM) (β = 0.287; p < 0.001), reinforcing the importance of regulatory clarity and institutional incentives in the adoption of sustainable practices.

Hypothesis 7 (H7) showed the strongest association in the model, demonstrating that implementation (IM) is significantly associated with results (REs) (β = 0.751; p < 0.001). This result highlights that the implementation of reuse technologies is linked to tangible benefits, such as resource savings and stakeholder recognition.

Hypothesis 2 (H2) showed a positive and significant association between organisational culture (OC) and implementation (IM) (β = 0.255; p = 0.013), confirming that values and environmental responsibility within the organisation are favourably associated with the adoption of reuse technologies.

In contrast, the hypotheses related to direct associations of knowledge (KN), as well as the links between EP, IA, IF, OC, and OR with results (REs), did not reach statistical significance (p > 0.05) and were therefore rejected.

Hypothesis 1 (H1) was confirmed, showing that organisational resources (ORs) have a positive and significant association with implementation (IM) (β = 0.273; p = 0.001), confirming that the availability of financial resources and senior management commitment are linked to the adoption of water reuse technologies.

3.4. Indirect Associations Analysis (Mediation Pathways)

Table 7. Testing indirect or second-order hypotheses.

Hypothesis		Path	p Values	Standard Deviation (STDEV)	Confidence Interval
					2.5%	97.5%
H12	EP → IM → RE	0.259	0.003	0.087	0.095	0.438
H10	IA → IM → RE	−0.263	0.000	0.075	−0.415	−0.121
H13	IF → IM → RE	0.216	0.000	0.059	0.116	0.346
H11	KN → IM → RE	0.054	0.528	0.086	−0.120	0.215
H9	OC → IM → RE	0.191	0.020	0.082	0.028	0.351
H8	OR → IM → RE	0.205	0.002	0.065	0.077	0.333

Note. Path = standardised indirect effect coefficient; p values derived from bootstrap resampling (5000 subsamples, bias-corrected and accelerated method); STDEV = standard deviation of the bootstrap distribution of indirect effects; 2.5%/97.5% = lower and upper bounds of the 95% bias-corrected bootstrap confidence interval.

presents the indirect associations of the structural model through the implementation (IM) variable. The results show that Hypothesis 12 (H12) was confirmed, demonstrating that external pressure (EP) has a positive and significant association with results (REs) through implementation (β = 0.259; p = 0.003). This indicates that regulatory and social pressure is linked to improvements in outcomes when the organisation implements reuse technologies.

Hypothesis 10 (H10) was also significant, showing a negative association between individual attitudes (IAs) and results (REs) operating through implementation (β = −0.263; p < 0.001). This finding suggests that individual attitudes are negatively associated with organisational achievements through their relationship with implementation.

Hypothesis 13 (H13) was supported, confirming that the institutional framework (IF) is positively associated with results (REs) through implementation (β = 0.216; p < 0.001).

Hypothesis 9 (H9) also obtained empirical support, showing that organisational culture (OC) has a positive and significant association with results (REs) operating through implementation (β = 0.191; p = 0.020), confirming the association of organisational environmental values with the achievement of outcomes.

Hypothesis 11 (H11) was not supported, as knowledge (KN) did not show a significant indirect association with outcomes (β = 0.054; p = 0.528).

Hypothesis 8 (H8) was supported, showing that organisational resources (ORs) are positively associated with results (REs) through implementation (β = 0.205; p = 0.002).

This suggests that the availability of financial and technical resources is linked to organisational benefits when it materialises in effective implementation processes.

4. Discussion

4.1. The Negative Association of Individual Attitudes with Technological Implementation

Among the findings of this study, the significant negative path between individual attitudes and the implementation of water reuse technologies (β = −0.350; p < 0.001) stands out as a particularly unexpected result that warrants careful interpretation. This pattern appears to challenge long-established theoretical propositions that have generally positioned favourable attitudes as facilitators of technology adoption. Our results, however, are broadly consistent with an emerging body of evidence documenting an “attitude-behaviour gap” in organisational settings, wherein positive individual perceptions do not necessarily translate into effective implementation, especially where structural constraints and institutional power dynamics exert a conditioning influence [9,12,20]. It should be noted that these findings must be interpreted within the boundaries of a cross-sectional design, which precludes definitive causal inference, and that replication across different institutional configurations would be necessary to ascertain the robustness of this association.

Theoretical Interpretation of the Paradoxical Effect of Attitudes: The negative association identified aligns with emerging scholarship on “implementation gaps” [9,12], wherein favourable attitudes that are disconnected from available structural capacities may generate cycles of frustration and organisational resistance. In settings characterised by limited resources, positive individual dispositions can function as markers of aspirational expectations which, upon encountering budgetary or institutional restrictions, may transform into disengagement from initiatives perceived as unattainable. To probe this unexpected pattern more rigorously, supplementary analyses were performed. First, the zero-order Pearson correlation between IA and IM was examined, yielding a moderate positive coefficient (r = 0.412, p < 0.001). This indicates that the simple bivariate association between individual attitudes and implementation is positive prior to the inclusion of additional predictors in the model. This divergence between the positive bivariate relationship and the negative structural path coefficient (β = −0.350) constitutes evidence that is suggestive of a classical statistical suppression effect, as described in the methodological literature [33,37]. Specifically, when external pressure (EP) and institutional framework (IF) are incorporated into the structural model, these variables appear to absorb a substantial portion of the shared variance between IA and IM. The residual attitudinal variance that remains after partialling out institutional influences exhibits an inverse association with implementation, a pattern that is consistent with—though does not conclusively establish—suppression dynamics as characterised by MacKinnon et al. (2000) [58] and Cenfetelli and Bassellier (2009) [59]. As a robustness check, the model was re-estimated excluding EP, which yielded a non-significant IA → IM coefficient (β = 0.087, p = 0.312), lending additional support to the interpretation that the negative path emerges specifically when institutional variables absorb shared variance. These findings should be interpreted with appropriate caution given the cross-sectional nature of the data. The evidence is consistent with the possibility that IA items capture symbolic endorsement of environmental technologies—reflecting socially desirable pro-environmental discourse—rather than genuine operational commitment. Nevertheless, alternative explanations cannot be excluded, and future research employing longitudinal designs would be necessary to confirm the directionality and mechanism underlying this association. The pattern may also reflect specificities of the Peruvian institutional context, where the gap between environmental discourse and organisational capacity has been documented in prior research [9,20], and generalisation to other settings should be undertaken cautiously.

4.2. The Centrality of Implementation as a Total Mediating Variable

A key finding of the structural model is that all associations with outcomes are channelled exclusively through the implementation variable, forming a pattern of total mediation that theoretically validates the critical importance of translating capabilities into concrete actions. This result corroborates systemic perspectives on environmental innovation that emphasise the inadequacy of resources, regulatory frameworks or organisational cultures when they do not materialise in specific operational practices [22,24,26].

The observed mediation pattern, where associations with outcomes are mainly channelled through implementation, may reflect the sequential nature of technology adoption processes, where resources, institutional frameworks and psychosocial factors need to be translated into concrete actions before generating measurable benefits. However, the absence of significant direct associations with outcomes may also indicate limitations in the model specification or the need for additional mediating variables not considered in this study [28,37].

4.3. Hierarchy of Structural Factors and Identified Association Patterns

The results reveal a clear hierarchy of determining factors, with external pressure showing the strongest association with implementation (β = 0.345), followed by the institutional framework (β = 0.287), organisational resources (β = 0.273) and organisational culture (β = 0.255). This associational pattern is consistent with institutionalist perspectives that position regulatory and social pressures as significant predictors of organisational environmental innovation within the model framework [12,21,22].

The primacy of external pressure is consistent with theoretical approaches to institutional configurations and environmental governance developed in the conceptual framework. In contrast to theoretical frameworks that privilege internal factors such as organisational culture or leadership, our findings suggest that in the Latin American contexts studied, exogenous forces are more strongly associated with implementation than endogenous capacities, supporting the relevance of regulatory frameworks in the adoption of circular practices [13].

In particular, the significance of the institutional framework as the second most relevant factor confirms the limitations of institutional fragmentation and regulatory gaps as systemic barriers to the implementation of the circular economy [12]. This hierarchy implies that public policy interventions should prioritise the strengthening of clear regulatory frameworks and economic incentive systems over technical training or cultural awareness programmes.

4.4. The Non-Significance of Technical Knowledge in Technology Adoption

Contrary to theoretical expectations derived from the conceptual framework on psychosocial dimensions, technical knowledge did not show significant associations with implementation (β = 0.071; p = 0.525). This finding challenges assumptions implicit in the literature on environmental management that position technical expertise as a fundamental prerequisite for the adoption of complex technologies.

The absence of significant associations with specialised technical knowledge may reflect a “ceiling effect” where the restricted distribution of expertise in circular water economy in the context studied limits the statistical variability necessary to detect effects [6,10]. Alternatively, it may indicate that in resource-constrained contexts, technical knowledge operates as an enabling prerequisite rather than a differentiating driver.

On the other hand, the insignificance of knowledge may reflect specificities of the Peruvian context, where access to specialised technical expertise in circular water economy remains limited, generating restricted distributions that reduce the statistical variability necessary to detect significant effects. Additionally, it is plausible that in contexts of limited resources, organisations prioritise the implementation of proven and accessible technologies over technically sophisticated solutions that require specialised expertise [9,10].

4.5. Theoretical and Methodological Contributions to the Field of Circular Economy

This study constitutes the first application of multidimensional SEM to analyse factors conditioning the implementation of the circular water economy in Latin American contexts, overcoming the limitations of one-dimensional approaches identified in the theoretical framework. The model’s high explanatory power (R² = 67.7% for implementation; R² = 82.3% for results) demonstrates the theoretical relevance of integrative frameworks that recognise the multi-causal nature of organisational environmental innovation phenomena.

Consistent with the conceptual developments presented, the findings validate the need for systemic approaches that transcend simplistic dichotomies between technical and social factors, revealing complex configurations where organisational, psychosocial, and institutional variables interact in a non-linear manner to produce differentiated outcomes. In contrast to previous studies that have examined isolated factors, this model demonstrates that successful implementation requires specific alignments between multiple levels of analysis, confirming the limitations of fragmented approaches [21,24].

4.6. Implications for Public Policy and Organisational Management

Actionable Pathways for Circular Water Economy Acceleration. The findings reveal a clear message for policymakers and managers: institutional factors matter more than internal capabilities. This challenges conventional approaches that emphasise training and awareness, suggesting instead that effective interventions must reshape the institutional environment in which organisations operate. Priority Public Policy Interventions. Our results demonstrate that external pressure (β = 0.345) and institutional framework (β = 0.287) exert the strongest influence on adoption. This means governments should prioritise four concrete actions: First, regulatory clarity is essential. Peru’s 2015 Framework Law remains largely unimplemented because it lacks operational regulations specifying water quality standards for different reuse applications, clear permitting procedures with explicit timelines, and defined liability frameworks [3,8].

Without these details, organisations face paralysing uncertainty about compliance requirements. Second, economic incentives must create compelling business cases. Effective incentive design includes progressive tariffs that penalise excessive freshwater consumption, capital subsidies covering 30–50% of treatment infrastructure costs, accelerated depreciation allowances for circular economy investments, and preferential access to green financing [9,13,14]. The modest effect of organisational resources (β = 0.273) indicates that many organisations cannot self-finance these transitions without external support. Third, institutional fragmentation must be addressed. Our study documented 14 overlapping agencies in Cajamarca alone, each claiming jurisdiction over water management aspects. Single-window permitting authorities can dramatically reduce transaction costs and regulatory uncertainty [3,8,12]. Fourth, demonstration programmes in high-visibility sectors can trigger mimetic adoption. Since external pressure reflects competitive and normative forces, strategically designed demonstration facilities generate peer learning and competitive repositioning that accelerate sector-wide diffusion [13,22].

Organisational Management Strategies. The total mediation pattern—where all factors influence outcomes exclusively through implementation (R² = 0.677)—carries a crucial implication: good intentions and adequate resources are worthless without systematic implementation capacity. Organisations must therefore prioritise: dedicated implementation structures that combine technical expertise, financial management, and regulatory compliance capabilities. The strong implementation-outcome link (β = 0.751) shows that organisations successfully translating plans into operations achieve substantial benefits regardless of initial resource constraints [28,37]. Phased technology deployment starts with low-complexity applications like cooling water recirculation before advancing to sophisticated wastewater treatment systems.

This staged approach prevents the frustration documented in our negative attitude-implementation finding (β = −0.350), building confidence through early wins [20]. Strategic alliances with technology providers and peer organisations to access scarce expertise. The non-significance of knowledge (β = 0.071) suggests widespread expertise deficits that individual organisations cannot remedy alone [6,10]. Cultural embedding through formal performance evaluation and recognition programmes. While organisational culture showed modest effects (β = 0.255), this reflects existing cultures rather than change potential deliberate interventions can strengthen normative support [9,20,25]. Context-specific adaptation implementation strategies must account for sectoral and regional variations.

Manufacturing facilities (42% of our sample) should prioritise closed-loop process water systems where technical specifications are well-established. Public services and construction sectors (47% of sample) face greater resource constraints but can achieve rapid implementation in standardised non-potable applications like landscape irrigation or vehicle washing. Regionally, Lima’s high water stress and developed infrastructure favour economic incentives and regulatory enforcement. Trujillo’s agro-industrial concentration enables sectoral coordination and technical standardisation, as facilities share similar wastewater characteristics. Cajamarca requires institutional consolidation before technical deployment, given the documented fragmentation across 14 agencies [3,8]. Finally, effective policy requires monitoring systems tracking regulatory implementation rates, economic incentive utilisation, organisational adoption rates, and achieved outcomes. This enables adaptive refinement rather than assuming interventions work as designed [8,13].

4.7. Specific Contextual Configuration

The patterns of association identified emerge within a specific institutional, economic and cultural configuration of the Peruvian context studied, characterised by developing regulatory frameworks, limited institutional capacities and asymmetries in access to advanced technologies. This contextual configuration fundamentally conditions the dynamics of technological adoption observed, suggesting that the primacy of external factors over internal factors may reflect specific characteristics of contexts where state institutional capacity operates as a critical limiting factor.

4.8. Theoretical Contributions: Integrating Empirical Findings with Resource-Based View, Theory of Planned Behaviour and Institutional Theory

The empirical findings enable critical examination of how established theoretical frameworks explain circular water economy adoption patterns in emerging economy contexts, revealing both convergences and divergences from theoretical expectations. Institutional Theory and the Primacy of External Factors. The finding that external pressure shows the strongest association with implementation (β = 0.345, p < 0.001), followed by institutional framework (β = 0.287, p < 0.001), is consistent with institutional theory’s core proposition that organisational behaviour is substantially linked to exogenous institutional forces [12,22]. This pattern corroborates DiMaggio and Powell’s institutional isomorphism framework, suggesting that coercive mechanisms (regulatory pressures) and normative mechanisms (stakeholder expectations) are more strongly associated with implementation than organisational resources or cultural predispositions in contexts characterised by developing regulatory frameworks and limited institutional capacity [22].

However, our findings also reveal important boundary conditions for institutional theory. The identified hierarchy—where external pressure exceeds institutional framework effects—suggests that in fragmented regulatory environments, informal pressures (stakeholder expectations, competitive mimicry) may compensate for weak formal institutions. This challenges the assumption that formal regulatory clarity is the primary institutional mechanism, demonstrating that normative and mimetic forces can operate independently of codified frameworks [12,13].

4.9. Resource-Based View: Conditional Relevance of Organisational Capabilities

The significant positive associations of organisational resources (β = 0.273, p = 0.001) and organisational culture (β = 0.255, p = 0.013) with implementation provide partial support for resource-based view propositions regarding the role of heterogeneous capabilities in technology adoption [9,24]. However, the relative magnitude of these effects—substantially weaker than institutional factors—challenges RBV’s traditional emphasis on internal capabilities as primary sources of competitive advantage in environmental innovation [24]. This pattern suggests an important theoretical refinement: in contexts where institutional pressures are strong and resources are broadly constrained, organisational capabilities operate as enabling conditions rather than differentiating drivers. RBV may retain greater explanatory power in institutional environments characterised by regulatory stability and resource heterogeneity, but exhibits limited predictive capacity in emerging economies where external pressures dominate and resource scarcity is widespread [9,10]. This finding extends Chowdhury et al.’s (2022) [9]. observation that organisational factors exhibit context-dependent effects, demonstrating that institutional configuration moderates the relevance of resource-based determinants.

4.10. Theory of Planned Behaviour: Paradoxical Effects and Theoretical Limitations

The counterintuitive negative association between individual attitudes and implementation (β = −0.350, p < 0.001) represents a fundamental challenge to theory of planned behaviour’s core assumption that favourable attitudes facilitate behavioural adoption [19,20]. It is important to acknowledge a contextual distinction when interpreting this result. TPB was originally formulated to explain individual decision-making, where the person holding the attitude is the same agent who executes the behaviour. In the present study, however, individual attitudes capture personal perceptions of decision-makers, while implementation reflects a collectively determined organisational process that depends on budgets, infrastructure, cross-departmental coordination, and institutional authorisation. This cross-level asymmetry—individual cognition predicting organisational action—may partly account for the paradoxical relationship observed, independently of TPB’s internal validity at the individual level. Rather than disconfirming TPB as a theory, our findings suggest that its direct application to organisational-level outcomes requires careful adaptation that accounts for the structural distance between individual disposition and collective execution. Emerging extensions of TPB addressing “implementation gaps” provide theoretical grounding for this paradox [20]. When individuals possess favourable attitudes but perceive insurmountable structural barriers (resource constraints, institutional fragmentation, and technical complexity), cognitive dissonance may generate frustration and resistance rather than facilitation [12,20]. In organisational contexts characterised by limited agency, positive attitudes disconnected from perceived behavioural control may produce counterproductive psychological dynamics. The non-significance of knowledge (β = 0.071, p = 0.525) similarly challenges TPB’s specification that perceived behavioural control partially derived from actual competence—facilitates behaviour [19,20].

This null finding may reflect either ceiling effects (uniformly limited expertise across the sample) or, more fundamentally, that technical knowledge operates as a necessary but insufficient prerequisite rather than a differentiating determinant in resource-constrained contexts [6,10]. Theoretical Integration: Towards Context-Sensitive Multi-Level Frameworks. Collectively, these findings demonstrate that no single theoretical perspective adequately explains circular water economy adoption patterns. Institutional theory provides the strongest explanatory power for primary drivers, RBV explains variance in implementation capacity conditional on institutional pressures, and TPB exhibits limited applicability in contexts where individual agency is constrained by structural factors.

This pattern validates calls for integrative theoretical frameworks that recognise context-dependent activation of different causal mechanisms [9,13,24,26]. Importantly, the total mediation pattern where all associations with outcomes operate through implementation suggests that theoretical frameworks focused exclusively on adoption intentions or capability development may be insufficient. Effective theories must incorporate implementation processes as distinct constructs rather than assuming automatic translation from resources, attitudes or institutional support to organisational outcomes [28,37]. This finding extends institutional and resource-based theories by demonstrating that successful circular economy transitions require explicit attention to implementation mechanisms, not merely favourable antecedent conditions.

4.11. Limitations and Directions for Future Research

This study has fundamental methodological limitations that condition the interpretation of findings. The cross-sectional design prevents the establishment of temporal directionality and definitive causal relationships, while non-probabilistic sampling restricts generalisation beyond the specific organisations studied, particularly considering the institutional heterogeneity that characterises Peruvian contexts. Several additional measurement limitations warrant acknowledgement. First, all constructs were measured through self-report questionnaires administered to a single respondent per organisation, which introduces potential self-report bias and social desirability effects, particularly for items assessing pro-environmental attitudes and implementation achievements [35,55]. The absence of objective implementation data (e.g., documented water reuse volumes, infrastructure audits, or third-party verification) means that the implementation and results constructs reflect perceived rather than verified organisational performance. Second, the unit of analysis requires clarification: while constructs are theoretically specified at the organisational level, the measurement captures individual perceptions of organisational decision-makers, creating a potential gap between reported and actual organisational conditions [9,20]. Third, external validity is limited; the findings should not be generalised beyond the specific population of decision-makers in the three regions studied without independent replication.

Critical Contextual Delimitation: The sample is limited to three Peruvian regions selected for operational convenience, not for national or regional statistical representativeness. Lima (urban water stress), Trujillo (export-oriented agribusiness) and Cajamarca (mining with socio-environmental conflicts) constitute specific contexts whose particular hydroeconomic, institutional and cultural characteristics fundamentally condition the patterns of association identified. Extrapolation to other national, regional or sectoral contexts lacks empirical basis and would constitute a methodologically unjustified overgeneralisation.

The exceptional coefficients of determination (R² = 67.7% for implementation; R² = 82.3% for outcomes) constitute a statistical anomaly in organisational research that may reflect: (1) successful specification of determining factors, (2) undetected residual multicollinearity, (3) common method bias derived from the single survey design, or (4) model overfitting. High HTMT correlations between constructs (EP–IA: 0.880; EP–IF: 0.860) suggest conceptual redundancy that compromises discriminant validity, indicating a need for theoretical refinement. The total absence of direct associations with outcomes constitutes a methodologically suspect pattern of complete mediation that may indicate omission of relevant variables or incorrect model specification. These exceptionally high values require cautious interpretation and replication in independent samples before confirming the robustness of the model.

Future research should employ: (1) longitudinal designs that allow for legitimate causal inference, (2) comparative studies across diverse institutional contexts, (3) mixed methodological approaches that integrate quantitative analysis with organisational case studies, (4) cross-validation with independent samples, and (5) inclusion of additional mediating variables (organisational absorptive capacity, resistance to change) and contextual moderating effects to develop more sophisticated theories on circular water economy in emerging economies.

5. Conclusions

It is important to note that the association patterns identified emerge specifically within the context of three Peruvian regions (Lima, Trujillo, and Cajamarca) with particular hydroeconomic, institutional, and cultural characteristics that fundamentally condition the dynamics observed. Lima exhibits characteristics of urban water stress, Trujillo corresponds to an agro-industrial export context, and Cajamarca reflects a mining environment with socio-environmental tensions. This specific contextual configuration limits direct extrapolation to other national, regional or sectoral contexts, which lacks empirical basis without additional validation. Replication through longitudinal designs in diverse institutional contexts is a critical priority for validating the transferability of these findings.

Within these contextual limitations, this study suggests that in the Peruvian regions studied, the adoption of circular water economy technologies operates through complex configurations that challenge traditional theoretical frameworks developed in more consolidated institutional contexts. The findings reveal that external structural factors are more strongly associated with technological adoption than internal organisational capacities. The empirical hierarchy, in which external pressure outweighs organisational resources, institutional frameworks and organisational culture, contradicts decades of research that privileges endogenous factors, suggesting that, specifically in contexts of limited institutional development such as those studied, organisations respond primarily to external incentives. This configuration may reflect particular characteristics of emerging economies where state institutional capacity operates as a critical limiting factor.

The total mediation identified reveals that all organisational, psychosocial and institutional factors need to be translated into actions and implementations in order to generate measurable results. This configuration theoretically validates that resources, regulatory frameworks or organisational cultures are irrelevant if they do not materialise into functional operating systems, positioning implementation as a critical link between latent capabilities and observable results.

The negative association between favourable individual attitudes and implementation is the most disruptive finding, challenging fundamental assumptions in environmental psychology. This result suggests that positive attitudes disconnected from structural capabilities are associated with implementation resistance, revealing a pattern where favourable perceptions can paradoxically operate as barriers when not backed by adequate organisational support.

Methodologically, this research sets critical precedents as the first application of multidimensional SEM to examine the circular water economy by integrating organisational, psychosocial, and institutional levels. Its exceptional explanatory power exceeds typical standards in organisational research, demonstrating the potential of integrative approaches that transcend disciplinary fragmentation.

The findings suggest possible directions for public policy in contexts similar to the one studied. The empirical primacy of external pressure indicates that more effective government interventions operate through regulations that generate clear sectoral expectations rather than training programmes. For organisations, total mediation indicates that investments in implementation capabilities are the most direct route to benefits, regardless of initial constraints.

This study provides pioneering empirical evidence on factor configurations associated with technology adoption in contexts characterised by specific institutional constraints, establishing theoretical foundations for advancing towards a more sophisticated understanding of organisational transitions towards sustainability in developing country contexts.