Resource Endowments, Value Cognition, and Strategic Risk-Taking: Explaining Carbon-Reduction Investments in Port Enterprises

Abstract

Against the backdrop of decarbonization in global maritime transport and logistics systems, port enterprises play a role in enhancing sustainable transport efficiency and system optimization through investments in carbon-reduction technologies. Situated within the institutional context of China’s “Dual-Carbon” targets, this study integrates Conservation of Resources theory and Behavioral Decision Theory to develop a dual-path analytical framework explaining carbon-reduction technology investment in port enterprises. A three-stage mixed-methods design is employed. First, grounded theory identifies four key resource categories: individual, conditional, material, and energy resources. Second, based on structural equation modeling and conditional process analysis of survey data from 372 port enterprise managers, the results show that individual and conditional resources significantly promote technology investment by enhancing perceived utility, while material resources exert a positive effect by increasing risk preference; the effect of energy resources is not significant. Environmental strategic orientation strengthens these relationships, whereas short-term performance pressure weakens them. Third, fuzzy-set qualitative comparative analysis reveals that multiple resource configurations can equivalently drive high levels of technology investment. Overall, this study uncovers the resource foundations and psychological mechanisms underlying carbon-reduction technology investment in port enterprises, offering empirical evidence for green technology investment in sustainable maritime transport and logistics.

1. Introduction

The intensification of global climate change has made green and low-carbon development a core issue within global governance systems [1]. In China, ports serve as critical hubs in international shipping networks and integrated logistics systems, yet they are characterized by high energy consumption and high carbon emissions, thereby facing substantial decarbonization pressure under the national “Dual-Carbon” strategy [2]. The “Dual-Carbon” strategy refers to China’s goals of achieving peak carbon emissions before 2030 and carbon neutrality by 2060 [3], the substantive objectives of which are highly aligned with international agendas on “net-zero emissions” and “decarbonization.” In recent years, China’s coastal ports have continuously increased investments in shore power systems, intelligent energy-consumption monitoring, electrified cargo-handling equipment, and green propulsion technologies [4]; however, the overall level of implementation remains markedly uneven [5]. According to Fujian Daily, among the 327 non-oil chemical terminal berths in Fujian Province, the coverage rate of shore power facilities reached 69% in 2024. In 2023, approximately 66,000 vessels utilized shore power while berthing, with total electricity consumption of about 7 million kWh—representing an average annual growth of roughly 40.7% since 2020—yet considerable gaps remain in terms of comprehensive deployment and deep utilization [6]. Concurrently, surveys of ports in the Bohai Rim region reveal substantial disparities in shore power utilization rates: Qingdao Port and Rizhao Port have approached or reached full utilization, Qinhuangdao Port exceeds 90%, whereas some shore power berths at Jinzhou Port have long remained underutilized [7]. Consequently, despite the increasing technological maturity of port carbon-reduction solutions [8], significant heterogeneity persists across ports with respect to the intensity of low-carbon technology investment and actual implementation effectiveness [5]. This situation underscores the urgent need to examine the key driving mechanisms and institutional constraints shaping port decarbonization investment decisions from the perspectives of resource endowments and institutional context.

Existing studies have primarily explained firms’ low-carbon investment behavior from the perspectives of policy pressure [9,10], green regulatory environments [11,12], and techno-economic considerations [13,14]. However, they have relatively neglected the central role of firms’ and managers’ internal resource endowments and their dynamic evolution in shaping investment decisions. For port enterprises, resource foundations not only determine the objective feasibility of low-carbon technology adoption but also profoundly influence managers’ subjective evaluations of expected returns, costs, and uncertainty [15]. Nevertheless, existing research lacks a systematic explanation of resource heterogeneity, resource-threat contexts, and dynamic resource mechanisms, making it difficult to fully capture the actual decision-making logic underlying green transformation in port enterprises. Meanwhile, carbon-reduction technologies are characterized by high costs, high uncertainty, and strong strategic attributes, rendering related investments inherently complex behavioral decision-making processes [16]. As a typical heavy-asset and policy-sensitive industry, port enterprises expose managers to heightened cognitive pressure and environmental complexity [17]; yet prior studies have rarely examined these issues from a psychological-mechanism perspective. Moreover, under multiple contextual constraints such as policy targets, market pressure, and short-term performance demands [18,19], how contextual factors moderate the effects of resources and psychological mechanisms on low-carbon investment decisions remains insufficiently explored.

Given the resource heterogeneity and decision-making complexity exhibited by port enterprises in carbon-reduction technology investment, this study first collects primary data through in-depth interviews and employs the three-stage coding procedure of grounded theory to identify key resource categories emphasized in port enterprises’ investment decisions. The inductively derived resource types are then systematically compared with the resource categories proposed in Conservation of Resources (COR) theory to assess the applicability of this theoretical framework within the port context. Building on this foundation, and drawing on Behavioral Decision Theory (BDT), the study follows a “resources–psychological mechanisms–investment behavior” analytical logic to develop two distinct pathways—the value cognition pathway and the strategic risk-taking pathway—to explain how different types of resources influence port enterprises’ low-carbon technology investment through subjective utility evaluation and risk preference. Subsequently, covariance-based structural equation modeling (CB-SEM) and PROCESS are employed to test the proposed linear and conditional mechanisms, while fuzzy-set qualitative comparative analysis (fsQCA) is further applied to identify multiple equivalent resource configurations that drive high levels of technology investment. This mixed-methods approach enables a more comprehensive understanding of the complexity underlying green investment decisions in port enterprises. This study seeks to address the following core research questions:

RQ1. What key resources are involved in port enterprises’ carbon-reduction technology investment, and do these resources support the applicability of COR theory in this context?

RQ2. How do different types of resources influence managers’ low-carbon technology investment decisions, and how do environmental strategic orientation and short-term performance pressure function as contextual moderators?

RQ3. Are there multiple resource configurations that can jointly drive high levels of carbon-reduction technology investment?

This study makes significant contributions at both the theoretical and practical levels. Theoretically, by integrating COR theory with BDT, it develops a dual-path analytical framework of “resources–psychological mechanisms–investment behavior,” elucidating how different types of resources influence port enterprises’ carbon-reduction technology investment through two distinct psychological mechanisms: value cognition and strategic risk-taking. In doing so, the study extends the explanatory boundaries of research on corporate green investment behavior. Moreover, by combining grounded theory, CB-SEM, PROCESS, and fsQCA, this study captures the complex structure of low-carbon technology investment in ports from three complementary perspectives—qualitative mechanisms, linear effects, and configurational pathways—thereby offering an integrated multi-method analytical paradigm for green technology adoption research. From a practical perspective, the findings provide empirical insights for port enterprises seeking to optimize resource allocation and enhance the efficiency of low-carbon technology investment, while also informing policymakers in the design of differentiated policy instruments and supporting managerial decision-making and training initiatives in the port sector.

The remainder of this paper is organized as follows. Section 2 systematically reviews the core concepts and relevant literature on COR theory and BDT, clarifying the theoretical foundations and integrative logic of the study. Section 3 draws on interview data and applies the three-stage coding procedure of grounded theory to identify key resource types involved in port enterprises’ carbon-reduction technology investment decisions. Section 4 develops the research model and hypotheses. Section 5 describes the research methods and analytical procedures. Section 6 reports the results of CB-SEM, PROCESS mediation and moderation analyses, and fsQCA. Section 7 discusses the main findings, derives theoretical implications and managerial insights, and summarizes the study’s limitations and directions for future research. Section 8 concludes the paper by providing an overall response to the research questions.

2. Literature Review

In this study, RQ1 focuses on the key resources involved in port enterprises’ investments in carbon-reduction technologies and their theoretical applicability, corresponding to the core propositions of COR theory regarding resource types and the processes of resource acquisition and preservation [20]. RQ2 examines how resources influence technology investment through two psychological pathways—perceived value and strategic risk-taking. Specifically, COR theory explains how resource endowments shape managers’ resource security boundaries and loss sensitivity, while BDT captures managers’ trade-offs between subjective utility and risk preference under conditions of high uncertainty [21]. RQ3 further explores the equifinal driving mechanisms of multiple resource configurations, echoing the notions of resource accumulation and gain spirals as well as causal complexity, thereby providing a theoretical basis for employing fsQCA to reveal multiple pathways to technology investment.

2.1. Conservation of Resources (COR) Theory

COR theory, originally proposed by Hobfoll [22], is used to explain how individuals and organizations acquire, maintain, and allocate resources under conditions of stress and uncertainty. The theory posits that resources constitute the fundamental basis for organizational stability and goal attainment, and that individuals and organizations, when confronted with environmental change or potential threats, tend to maximize resource possession while minimizing resource loss [20]. COR classifies resources into multiple categories, including material resources (e.g., financial capital, equipment, and technological capabilities), energy resources (e.g., time and emotional energy), conditional resources (e.g., institutional arrangements, policy support, and organizational culture), and personal resources (e.g., self-efficacy, psychological resilience, professional skills, and managers’ strategic judgment capabilities) [23]. These resources interact with one another and jointly shape organizational decision-making and adaptive capacity in contexts of transformation, innovation, and external pressure [24]. Moreover, COR theory emphasizes resource spiral mechanisms, whereby resource-rich organizations are more likely to generate self-reinforcing cycles of sustained investment and resource accumulation, whereas resource-constrained organizations may fall into downward spirals of continuous resource depletion under stressful conditions [25]. Accordingly, firms’ resource endowments not only influence their short-term strategic choices but also profoundly shape long-term resource accumulation trajectories and competitive advantages, a dynamic that is particularly salient in the contexts of green transformation, digital upgrading, and technological change [26].

Compared with research perspectives that focus on a single type of resource, COR theory, through its multi-dimensional resource framework, provides a more operational and explanatory theoretical foundation and has been widely applied in fields such as management studies [27], strategic behavior [28], and organizational decision-making and innovation management [29]. However, most existing studies remain concentrated on the direct relationships between resources and behavioral outcomes [14,30], with limited attention to the psychological decision-making mechanisms through which resources influence high-risk and long-term investment behaviors. Particularly in the context of port enterprises—characterized by heavy asset intensity, strong regulation, and high uncertainty—resource endowments not only determine whether firms have the capacity to invest, but also more profoundly shape managers’ subjective evaluations of technological value and potential risks [15]. Moreover, COR-based studies often underemphasize the psychological pathways through which resources are translated into decision preferences, thereby constraining their explanatory power for complex strategic investment behaviors. This limitation suggests that reliance on COR theory alone is insufficient to fully uncover the actual decision logic underlying low-carbon technology investment in port enterprises, and that it is necessary to incorporate a behavioral decision-making perspective capable of capturing managers’ value trade-offs and risk judgments.

2.2. Behavioral Decision Theory (BDT)

BDT emerged in response to the limited explanatory power of the traditional fully rational decision-making paradigm. The theory traces its origins to early work on psychological judgment by Edwards [31] and was subsequently systematized through the heuristics-and-biases framework proposed by Tversky and Kahneman [32], as well as prospect theory developed by Kai-Ineman and Tversky [33]. BDT emphasizes the prevalence of bounded rationality, cognitive biases, and subjective value evaluation mechanisms in real-world decision-making processes [34], thereby challenging the assumptions of complete information, stable preferences, and utility maximization embedded in expected utility theory [35]. Extensive research indicates that actual decision-making is constrained by limited cognitive resources, time pressure, and information-processing costs, leading decision-makers to adopt satisficing strategies under conditions of bounded rationality [36,37]. Building on this foundation, prospect theory further demonstrates key psychological characteristics such as reference dependence, loss aversion, and probability weighting, suggesting that individuals tend to be risk-averse in gain contexts but may exhibit risk-seeking behavior in loss contexts [33]. In complex or information-overloaded environments, decision-makers frequently rely on heuristic rules to simplify judgment, which can reduce cognitive burden while simultaneously giving rise to systematic biases [32]. Overall, BDT conceptualizes real-world decision-making as a dynamic process characterized by the interplay of bounded rationality, heuristics, and biases [21].

In recent years, BDT has been widely applied and further developed in areas such as organizational behavior [38], sustainable entrepreneurship [39], and consumer decision-making [40]. However, existing BDT-based research has largely focused on individual-level cognitive biases or variations in risk preference [41,42], with relatively limited systematic examination of how organizational-level resource conditions shape decision-makers’ psychological judgment frameworks. In the context of low-carbon technology investment, a decision domain that is highly dependent on organizational resource support, managers’ value perceptions and risk preferences do not emerge in isolation but are deeply embedded in the resource structures they control and perceive [15]. Consequently, discussing behavioral decision-making mechanisms without anchoring them in the underlying resource base remains insufficient to explain why firms facing similar external pressures often exhibit markedly divergent investment behaviors. This practical challenge highlights the necessity of a contextualized integration of BDT and COR theory, advancing a coherent “resources–psychological mechanisms–investment behavior” analytical chain to systematically uncover the internal mechanisms underlying port enterprises’ low-carbon technology investment decisions.

3. Identification of Resource Types in Port Enterprises

To further identify the key resource types involved in port enterprises’ carbon-reduction technology investment decisions, this study adopts the three-stage coding procedure of grounded theory to conduct a systematic analysis of interview data. Through this process, core resource-related concepts are inductively extracted and subsequently compared with the resource classification framework proposed by COR theory, in order to examine the applicability and explanatory power of existing theoretical constructs within the context of port enterprises’ low-carbon transition. This approach not only helps to avoid potential context misalignment that may arise from the direct application of pre-established theoretical categories, but also ensures—through bottom-up empirical induction—that the identified resource types accurately reflect the practical concerns of port enterprises operating in decision environments characterized by high capital intensity, stringent regulatory constraints, and elevated uncertainty. On this basis, the study achieves a refined contextual characterization of low-carbon technology investment decisions while maintaining theoretical coherence, thereby providing a solid empirical foundation and conceptual grounding for the subsequent development of an analytical model that integrates both contextual embeddedness and theoretical explanatory power.

3.1. Collection and Preparation of Qualitative Data

Grounded theory is a well-established qualitative research paradigm that is particularly suitable for research domains in which theoretical development remains underdeveloped and research questions have not yet been sufficiently explained. Its core value lies in the systematic and iterative analytical procedures through which concepts are derived from empirical materials, categories are constructed, and contextually grounded theories are ultimately generated [43]. Although grounded theory was originally designed to develop entirely new theoretical constructs based on empirical data, its methodological features are equally applicable to the examination, extension, and refinement of existing theories [44]. Through the structured three-stage coding process—open coding, axial coding, and selective coding—researchers are able to assess, with theoretical sensitivity, the degree of alignment between established theories and empirical evidence in a data-driven analytical process, thereby evaluating whether existing theoretical frameworks adequately explain behavioral mechanisms in specific contexts [45]. Moreover, the highly flexible analytical logic of grounded theory allows it to be widely applied to diverse data sources, including interview transcripts, observational notes, and various forms of archival or secondary texts [46]. Given these methodological advantages, this study adopts the three-stage coding procedure of grounded theory as a key methodological approach to examine, validate, and further develop the applicability of existing theories in a novel context.

In the research implementation stage, this study disseminated recruitment information through the “Xiaohongshu” platform to identify eligible interview participants. Respondents were required to be at least 18 years of age and to occupy positions at the strategic or investment decision-making level within port enterprises, including heads or deputy heads of key functional departments such as operations management, safety and environmental protection, and strategy and planning. This sampling criterion ensured that the collected data adequately reflected the judgments and evaluations of core organizational decision makers. The interviews were conducted via telephone and focused primarily on the resource allocation processes involved in port enterprises’ investment decisions related to emission-reduction technologies. In terms of research procedures, this study followed the fundamental grounded theory paradigm of “theoretical sampling–constant comparison–theoretical saturation” [43], advancing data collection and analysis in an iterative manner. Ultimately, a total of 23 valid interview transcripts were obtained (see

Table 1. Demographic Characteristics of the Interview Sample (n = 23).

Category	Item	Frequency
Gender Composition	Male	12
	Female	11
Age Distribution (years)	18–30	2
	31–40	6
	41–50	11
	Above 50	4
Educational Background	Associate degree or Below	3
	Bachelor’s Degree	11
	Master’s Degree or Above	9
Functional Department Category	Operations Management	4
	Safety and Environmental Protection	4
	Strategy and Planning	6
	Investment and Financial Decision-Making	6
	Technology and Engineering Management	3
Interview Duration (minutes)	20–25	6
	26–30	9
	31–35	8

for respondent information), and all participants provided written informed consent in electronic form. To ensure data integrity and credibility, all interviews were fully audio-recorded with participants’ informed consent. Following the interviews, the research team transcribed all audio recordings verbatim, producing approximately 95,000 words of raw textual data, which provided a solid empirical foundation for the subsequent three-stage grounded theory coding process.

3.2. Data Coding and Analysis

Following the completion of data collection, the research team implemented a pre-established analytical procedure in which approximately two-thirds of the interview transcripts were designated as the primary dataset for analysis, while the remaining one-third was reserved for subsequent theoretical saturation testing. NVivo 11 was used to manage and organize the textual data, and a systematic analysis was conducted in accordance with the three-stage coding procedure proposed by Strauss and Corbin [47]. First, during the open coding stage, the researchers conducted a line-by-line and paragraph-by-paragraph examination of the transcripts in relation to the research objectives. The data were carefully deconstructed to extract conceptual statements that accurately reflected participants’ lived experiences. Statements with similar meanings or consistent logical implications were merged, while fragmented information lacking analytical value was eliminated. Through this process, a total of 21 initial concepts were identified. Second, in the axial coding stage, the concepts generated in the previous phase were compared and aggregated to identify underlying relationships and distinctions. Semantically convergent concepts were reorganized and integrated, resulting in the development of nine more explanatory subcategories. Finally, in the selective coding stage, the researchers further synthesized these subcategories by considering the behavioral characteristics of port enterprises in the context of carbon-reduction technology investment. This process culminated in the identification of four overarching core categories that captured the overall structural framework of the data. To assess the robustness of the coding scheme, the research team subsequently employed the reserved one-third of the interview transcripts to conduct a theoretical saturation test. By repeating the established coding procedures and comparing the results with the existing conceptual framework, no new key concepts or categories emerged, nor did any data conflict with the established structure. These findings indicate that the category framework achieved theoretical saturation. The detailed coding results are presented in

Table A1. Coding Results.

Representative Raw Data	Initial Concepts	Subcategories	Core Categories
There are now many types of green technologies, and we still need to continuously learn about the specific principles of technologies such as shore power, energy consumption monitoring, and electrified equipment.	Green technology awareness	a1 Professional Capability	A1 Personal Resources
Environmental policies are updated very quickly, especially carbon emission requirements for ports. We must clearly understand the policies before deciding whether to invest.	Understanding of environmental policies
When designing technical solutions, I have to carefully calculate costs, efficiency, and returns, which requires certain evaluation capabilities.	Technological evaluation capability
All technology investments involve uncertainty. We have to bear a certain level of risk and cannot avoid pressure altogether.	Tolerance for uncertainty	a2 Psychological Resilience
Port transformation is a long-term process. Personally, I have confidence in the low-carbon direction; otherwise, projects would be difficult to advance.	Confidence in low-carbon transformation
Some internal procedures are overly complicated and significantly delay the approval of technology projects. It would be better if the processes were clearer.	Clarity of organizational processes	a3 Institutional Supportiveness	A2 Condition Resources
The government has explicit requirements regarding shore power usage and carbon emissions, and we must comply with these regulations.	Regulatory compliance environment
If emission reduction targets could be linked to performance evaluation, we would be more motivated to implement projects.	Alignment of incentive systems
Information on many low-carbon technologies is not transparent, making it difficult for us to understand their actual effectiveness.	Transparency of technological information	a4 Information Accessibility
Sometimes we need to consult external experts, such as equipment suppliers or third-party institutions.	Access to external consulting channels
The infrastructure for shore power systems is not sufficiently developed, which makes additional investment more difficult.	Shore power infrastructure base	a5 Low-Carbon Infrastructure Base	A3 Material Resources
The charging and dispatching conditions for electrified equipment are inadequate, which constrains the application of these technologies.	Conditions for electrified equipment
Our annual budget for technology investment is limited and needs to be adjusted flexibly according to circumstances.	Flexibility of investment budgets	a6 Financial Investment Capacity
If funds could be flexibly reallocated across projects, we could implement low-carbon pilot projects more quickly.	Flexibility in capital utilization
Some old equipment can still be retrofitted; it is not always necessary to replace everything with new equipment.	Retrofit adaptability of existing assets	a7 Asset Retrofitability
When carrying out technological upgrades, we need to coordinate resources from multiple parties, and this capability is crucial.	Capability for implementing technological upgrades
Our team has been in very good overall condition recently, with high energy levels, which has improved efficiency when advancing tasks.	Level of energy and vitality	a8 Personal Vitality	A4 Energy Resources
At present, we are concentrating most of our efforts on low-carbon projects, so progress has been relatively smooth.	Degree of energy focus
Whether departments can cooperate smoothly directly determines the speed of project implementation.	Cross-departmental coordination capability	a9 Organizational Momentum
If top management shows strong commitment, projects move forward quickly; otherwise, they tend to be delayed.	Top management commitment
Whether our team has strong execution capability is the key to successfully implementing low-carbon projects.	Team execution vitality

.

Building on the foregoing analysis, this study preliminarily identifies the key factors influencing port enterprises’ investments in emission-reduction technologies, which mainly encompass Personal Resources, Conditional Resources, Material Resources, and Energy Resources. A comparison with the core classification logic of COR theory reveals that the resource types identified in this study do not extend beyond the boundaries of the COR framework, indicating that COR’s resource categorization remains highly applicable in the specific context of ports’ carbon-reduction technology investment. Nevertheless, the specific mechanisms, influence logic, and critical pathways through which these resource elements operate in the process of ports’ investment in emission-reduction technologies remain insufficiently clear. Accordingly, there is a need for further systematic empirical research to disentangle the dynamic mechanisms through which different types of resources shape technology investment decisions.

4. Model Development and Hypothesis Formulation

4.1. Model Development

COR theory emphasizes that individuals and organizations tend to acquire, retain, and enhance their key resources, thereby generating either “resource gain spirals” or “resource loss spirals” [25]. Integrating the resource typology of COR theory with the results of the three-stage grounded theory coding, this study classifies the resources that port enterprises can mobilize in promoting investments in carbon-reduction technologies into four categories: Personal Resources, Conditional Resources, Material Resources, and Energy Resources. Together, these resources constitute the foundational endowments that enable firms to achieve green technological transformation. Moreover, drawing on BDT, this study posits that the availability of different types of resources not only influences decision makers’ cognitive evaluations of investments in carbon-reduction technologies but also shapes their behavioral preferences under conditions of uncertainty. Specifically, subjective utility reflects decision makers’ value cognition, expected returns, and feasibility assessments regarding a given technology [48], whereas risk preference captures their propensity to bear risk when confronted with technological uncertainty, investment costs, and potential returns [49]. Accordingly, this study conceptualizes subjective utility and risk preference as key mediating variables through which resources affect carbon-reduction technology investment behavior, and develops two distinct pathways—the “value cognition pathway” and the “strategic risk-taking pathway”—to elucidate the underlying psychological mechanisms through which different resources promote technological investment via resource gain spirals.

Building on this framework, the study further incorporates moderating variables to clarify the boundary conditions under which these mediating effects operate. Prior research suggests that a strong environmental strategic orientation can reinforce organizations’ willingness to allocate resources toward green technologies [50,51]. At the same time, the mechanism underlying resource loss spirals implies that external constraints and short-term performance pressure may deplete organizational resources and weaken firms’ propensity to engage in green investment [52,53]. Based on these theoretical considerations, environmental strategic orientation and short-term performance pressure are introduced as contextual moderators to examine their moderating effects within the mediating mechanisms.

In summary, this study develops a conceptual model of port enterprises’ investment in carbon-reduction technologies that incorporates four types of resources, two mediating variables, and two contextual moderators (see Figure 1). The model systematically illustrates how multiple resource types drive firms’ technological investment through psychological mechanisms, while also elucidating the strengthening or weakening roles played by contextual factors.

4.2. Hypothesis Development

4.2.1. The Value Cognition Pathway

In this study, personal resources refer to the cognitive abilities, psychological attributes, and professional competencies embedded within port enterprise managers that are mobilized during carbon-reduction technology investment decision-making. These resources directly influence the quality of information processing, the capacity for value judgment, and subjective assessments of uncertainty, thereby shaping managers’ perceived utility of low-carbon technology investment. Perceived utility denotes managers’ overall evaluative judgment formed on the basis of their subjective expectations regarding the benefits of carbon-reduction technologies, such as improvements in environmental performance, cost savings, regulatory compliance advantages, and the strengthening of competitive position. According to COR theory, personal resources such as self-efficacy and psychological resilience enhance individuals’ sense of control over environmental events, leading them to prioritize resource gains rather than avoid uncertainty in decision-making [22]. As an intrinsic and self-oriented resource category, personal resources are regarded as among the most fundamental determinants of value judgment and goal pursuit [27]. They not only increase confidence and resilience when facing challenges but also heighten sensitivity to potential positive outcomes, thereby elevating subjective evaluations of behavioral consequences [24]. Empirical evidence across diverse contexts supports this mechanism. For instance, self-efficacy has been shown to significantly enhance individuals’ perceived usefulness and subjective utility evaluations of technological systems or AI tools [54,55], while positive psychological capital strengthens perceived task and work value [56]. Taken together, it can be inferred that, within the context of this study, higher levels of personal resources among managers are associated with more favorable subjective utility judgments regarding the potential benefits of carbon-reduction technologies. Accordingly, the following hypothesis is proposed:

H1: 

Personal resources have a positive effect on subjective utility.

In this study, conditional resources refer to a set of contextual resources available to or mobilized by port enterprise managers within organizational and policy environments that provide institutional safeguards, organizational support, technological and informational infrastructure, professional services, and external policy incentives for carbon-reduction technology investment. According to COR theory, conditional resources reduce uncertainty and implementation costs in the decision-making and execution processes, thereby mitigating potential threats of resource loss and encouraging managers to adopt resource-gain–oriented judgments in their value trade-offs [22]. Existing research across various technological application contexts has substantiated this mechanism. For example, external support and facilitating conditions have been shown to significantly enhance individuals’ perceived utility of technology use outcomes and overall value assessments [57,58]. Similarly, institutional and resource-based organizational support, as critical forms of conditional resources, strengthens managers’ expectations regarding the benefits of adopting emerging technologies such as AI [59]. Taken together, it can be inferred that, in the context of carbon-reduction technology investment examined in this study, managers with more abundant conditional resources are more likely to form more favorable subjective utility evaluations of the economic and environmental benefits associated with carbon-reduction technologies. Accordingly, the following hypothesis is proposed:

H2: 

Conditional resources have a positive effect on subjective utility.

In the present study, technology investment refers to the resource allocation behaviors undertaken by port enterprise managers to promote the application of carbon-reduction technologies, including financial expenditures for technology pilots, deployment, and diffusion; procurement of equipment and infrastructure; spending on technological retrofitting; investments in personnel training; and engagement with external technical services. Subjective utility reflects managers’ positive expectations regarding the value of technology investment and serves as a critical psychological driver of resource allocation and technology investment decisions under complex conditions [48]. A substantial body of research consistently demonstrates that perceptions of technological utility constitute a key antecedent of technology adoption, continued use, and organizational-level technology investment behavior [30,45]. When managers expect a technology to deliver performance improvements or strategic benefits, both their investment intention and the intensity of resource commitment increase significantly [25]. Recent empirical studies further corroborate this mechanism across diverse technological contexts. For instance, perceived utility has been shown to significantly promote the deep application of AI systems and improvements in organizational decision efficiency [60], and to be positively associated with firms’ investment levels in digital tools and digital technologies [61]. Similarly, managers’ subjective utility evaluations of digital tools such as social media effectively predict their usage intensity and investment levels [30]. Taken together, it can be inferred that, in the context of carbon-reduction technology investment in port enterprises, higher levels of perceived utility are associated with stronger investment intentions and greater resource allocation efforts, thereby resulting in higher levels of technology investment. Accordingly, the following hypothesis is proposed:

H3: 

Subjective utility has a positive effect on technology investment.

In the present study, environmental strategic orientation is defined as the extent to which a firm, at the overall strategic level, emphasizes environmental sustainability, the application of green technologies, and low-carbon transformation, as well as its tendency to systematically integrate environmental objectives into organizational strategy and resource allocation. Although existing research has not yet reached a consensus regarding the specific moderating role of environmental strategic orientation in the relationships between perceived utility and technology investment or between risk preference and technology investment, accumulating empirical evidence suggests that environmentally oriented strategies can significantly strengthen the translation of technological investment into green outcomes. For example, a digital green strategic orientation has been shown to amplify the positive effect of technology deployment on firms’ green performance [62]. Similarly, in the context of green product and process innovation, higher levels of green strategic orientation magnify the positive impact of technological innovation investment on green performance [63]. Further studies indicate that environmental or green entrepreneurial orientation not only enhances firm performance but also reinforces managerial preferences for green innovation and technology investment under conditions of heightened environmental pressure [64,65]. Collectively, these findings suggest that when organizations strengthen their environmental orientation at the strategic level, managers’ opportunity recognition, risk-taking tendencies, and resource allocation decisions are more likely to be directed toward green technologies and green innovation. Accordingly, it can be inferred that, in the context of port enterprises, a higher level of environmental strategic orientation strengthens the extent to which managers’ technology investment decisions—formed on the basis of perceived utility or risk preference—are translated into actual investments in carbon-reduction technologies. Accordingly, the following hypotheses are proposed:

H4: 

Environmental strategic orientation positively moderates the relationship between subjective utility and technology investment, such that the positive effect of subjective utility on technology investment is stronger at higher levels of Environmental strategic orientation.

H5: 

Environmental strategic orientation positively moderates the relationship between risk preference and technology investment, such that the positive effect of risk preference on technology investment is stronger at higher levels of Environmental strategic orientation.

4.2.2. The Strategic Risk-Taking Pathway

In this study, material resources refer to the stock of tangible resources that port enterprises can allocate and mobilize during carbon-reduction technology investment, including financial slack, investment budgets, equipment and infrastructure conditions, and asset retrofitting capacity. These resources expand the organization’s safety boundary, reduce potential failure costs, and thereby enhance managers’ tolerance for risk associated with highly uncertain technological investments. Risk preference denotes managers’ subjective tolerance for uncertainty and their psychological inclination to favor technological options with higher uncertainty but greater potential returns when weighing risks against benefits. According to COR theory, abundant material resources alleviate decision-makers’ perceptions of resource vulnerability in critical decisions, making them more inclined to adopt resource-gain–oriented strategies under uncertain conditions [22]. When managers command greater material resources, the perceived constraints associated with potential failure are weakened and the safety boundary expands accordingly, increasing their willingness to undertake high-risk, high-potential-return technological investments [49]. Prior research has corroborated this mechanism across different contexts. For example, higher levels of organizational slack resources or financial flexibility are consistently associated with stronger risk-taking tendencies [66,67,68]. Taken together, it can be inferred that, within the context of this study, the more abundant the material resources under managers’ control, the higher their tolerance for and preference toward high-risk technological investment. Accordingly, the following hypothesis is proposed:

H6: 

Material resources have a positive effect on risk preference.

In this study, energy resources refer to the reserves of effort, attention, and emotional momentum that port enterprise managers or organizations can continuously allocate to decision-making, coordination, and execution activities when advancing investments in carbon-reduction technologies. Such resources determine an organization’s capacity for sustained engagement and execution intensity in complex tasks. According to COR theory, energy resources such as time and effort can themselves be invested in acquiring and accumulating other resources, thereby triggering resource gain cycles and strengthening individuals’ perceived control over their resource environment [22]. When managers possess higher levels of energy resources, their subjective sensitivity to potential losses is reduced, as they are more confident in compensating for losses through subsequent resource deployment, which in turn leads to higher risk preference [69]. Existing research suggests that risk preference is shaped not only by objective benefit–cost structures but is also significantly influenced by the psychological load perceived at the time of decision-making and the individual’s coping capacity [70]. In high-uncertainty and high-complexity technology investment contexts, whether managers have sufficient energy to engage in project evaluation, cross-departmental coordination, and risk management directly affects their subjective tolerance for uncertainty [71]. Related empirical evidence further supports this reasoning: flexibly deployable time and energy resources strengthen firms’ risk-taking propensity [72]; energy resources are systematically associated with green innovation strategies [24]; and financial knowledge, conceptualized as a form of “knowledge-based energy resource,” significantly enhances individuals’ risk-taking levels [73]. Taken together, energy resources enhance individuals’ or organizations’ ability to continuously “regenerate resources” in uncertain environments, thereby increasing their willingness to bear risk and their overall risk preference. Accordingly, the following hypothesis is proposed:

H7: 

Energy resources have a positive effect on risk preference.

In highly uncertain contexts such as technological upgrading and innovation investment, risk preference is widely regarded as a key individual-difference variable influencing organizations’ exploratory investments and innovation-related decisions [49]. Qi et al. [74] argue that risk preference constitutes a core antecedent of risk decision-making, whereby a higher level of risk tolerance strengthens willingness to take risks and motivates organizations to select innovation projects that entail greater risk but offer higher potential returns. A substantial body of research across diverse organizational and industrial contexts has confirmed the positive effect of risk preference on technology-related investment behaviors. For example, managers with higher risk tolerance are more likely to encourage greater organizational investment in uncertain domains such as human capital and advanced technologies [75,76], and individual risk preference significantly influences the allocation of assets toward high-risk investments and entrepreneurial ventures [77]. In the context of green transformation and technological R&D, increasing executives’ risk preference has similarly been shown to enhance R&D intensity and the scale of technology investment [78]. Taken together, it can be inferred that in the highly uncertain decision-making context of carbon-reduction technology investment in port enterprises, managers with higher risk preference are more inclined to allocate greater resources to carbon-reduction technologies. Accordingly, the following hypothesis is proposed:

H8: 

Risk preference has a positive effect on technology investment.

In this study, short-term performance pressure refers to the internal and external pressures perceived by port enterprise managers due to organizational emphasis on short-term operational outcomes. Such pressure drives managers to focus more heavily on short-cycle performance returns, while relatively downplaying long-term value creation, innovation experimentation, and strategic technology investment [52]. Prior research indicates that intense short-term performance pressure shifts managerial decision orientations toward low-risk, quick-return behaviors, thereby suppressing investment in technologies and innovative activities characterized by high uncertainty and long payoff horizons [79]. At the same time, performance pressure has been shown to significantly weaken managers’ risk-taking behavior in innovation and strategic investment decisions [53,80]. More recent empirical evidence further demonstrates that under high short-term performance pressure, management systematically curtails long-term technology investment and exploratory innovation activities [81]. Collectively, these findings suggest that short-term performance pressure not only shapes managers’ risk attitudes but also disrupts the translation of technology value perceptions and risk preferences into actual innovation decisions. Accordingly, it can be inferred that in the context of this study, short-term performance pressure weakens the positive effects of managers’ favorable psychological factors on carbon-reduction technology investment. When short-term performance pressure is high, even managers with strong perceived utility or elevated risk preference may substantially restrain their engagement in technology pilot initiatives and investment intensity. Accordingly, the following hypothesis is proposed:

H9: 

Short-term performance pressure negatively moderates the relationship between subjective utility and technology investment, such that the positive effect of subjective utility on technology investment is weaker at higher levels of short-term performance pressure.

H10: 

Short-term performance pressure negatively moderates the relationship between risk preference and technology investment, such that the positive effect of risk preference on technology investment is weaker at higher levels of short-term performance pressure.

The existing literature provides limited explicit theoretical or empirical evidence regarding the mediating roles of subjective utility and risk preference in the relationship between resource endowments and technology investment. However, based on the theoretical logic underlying Hypotheses H1–H3 and H6–H8 developed in this study, it can be inferred that subjective utility and risk preference are likely to serve as key mediating mechanisms within the value cognition pathway and the strategic risk-taking pathway, respectively. To further elucidate these potential mechanisms, this study will conduct exploratory mediation analyses in the data analysis stage. This approach is intended to deepen understanding of the proposed theoretical pathways and to offer insights that may inform future theoretical development and empirical research.

Building on the above hypothesis development and analytical logic, it is necessary to clarify the strengths and limitations of the overall analytical framework adopted in this study. By integrating COR theory and BDT, this research develops a dual-path model that systematically elucidates how resource endowments influence port enterprises’ carbon-reduction technology investment through perceived utility evaluation and risk preference under conditions of high uncertainty, thereby enhancing the process-oriented explanatory power of investment decision-making mechanisms. However, the model primarily focuses on two core psychological mechanisms—perceived utility and risk preference—and restricts external contextual factors to environmental strategic orientation and short-term performance pressure. As a result, other potential cognitive biases, interactions among different types of resources, and the influence of broader institutional environments are only partially captured. These mechanisms warrant further extension and validation across diverse contexts and research designs.

5. Research Methodology and Procedure

5.1. Questionnaire Design

All measurement scales used in this study were adapted from well-established instruments in the existing literature, with appropriate modifications. To enhance content validity, one expert each from the fields of behavioral science, port and shipping management, and carbon emissions and green technology was invited to conduct multiple rounds of review and revision. In addition, to ensure linguistic equivalence and cultural appropriateness of the measurement instruments in the context of mainland China, the scale development process strictly followed translation and back-translation procedures [82,83]. Specifically, two bilingual researchers with relevant research backgrounds independently translated the original English scales into Chinese. Subsequently, another bilingual scholar who was not involved in the initial translation performed a back-translation of the Chinese version, and the back-translated items were compared item by item with the original English statements. Any semantic deviations, pragmatic ambiguities, or potential cultural mismatches identified during this comparison were systematically discussed and repeatedly revised by the research team in consultation with the expert panel to ensure consistency in both semantic expression and conceptual meaning across all measurement items.

The final instrument comprised nine latent variables measured by a total of 31 items (see

Table A2. Measurement items.

Variable		Measurement Items	Source
Personal Resources (PR)	PR1	I am capable of assessing the feasibility and applicability of different carbon-reduction technologies.	Güler and Çetin [110] Contreras et al. [111]
	PR2	I am confident in promoting the implementation of green technologies in this port.
	PR3	When facing difficulties in green technology investment, I am able to identify effective solutions.
	PR4	I possess the professional knowledge required to understand and evaluate carbon emission issues in ports.
Material Resources(MR)	MR1	The port has sufficient budgetary resources to support investments in carbon-reduction technologies.	Li et al. [45]
	MR2	Existing equipment and infrastructure are adequate to meet the requirements for implementing green technologies.
	MR3	The port’s existing equipment has retrofit potential and can be adapted to emission-reduction technologies without full replacement.
Condition Resources(CR)	CR1	Top management of the port clearly supports investments in carbon-reduction technologies.	Alkandi et al. [112] Li et al. [45]
	CR2	The current policy environment (e.g., regulations and subsidies) facilitates the adoption of green technologies.
	CR3	We are able to obtain professional support from external technical institutions or suppliers.
	CR4	The port has the necessary human and time resources to advance green technology projects.
Energy Resources (ER)	ER1	I am generally energetic when dealing with decisions related to carbon-reduction technologies.	Szilvassy and Širok [113] Abuzaid et al. [114]
	ER2	I am able to maintain a high level of focus when evaluating green technology solutions.
	ER3	Active support from top management motivates us to advance low-carbon projects more effectively.
Subjective Utility (SU)	SU1	I believe that adopting carbon-reduction technologies will significantly enhance the port’s long-term competitive advantage.	An et al. [115] Hu et al. [45]
	SU2	I believe that green technologies can improve port operational efficiency and generate potential benefits.
	SU3	I believe that implementing carbon-reduction technologies will have a clearly positive effect on the port’s social reputation.
Risk Preference (RP)	RP1	When facing new green technologies, I am willing to accept a certain degree of uncertainty.	Zhang et al. [116] Setiawan et al. [117]
	RP2	Even when implementation risks exist, I am inclined to try new carbon-reduction technology solutions.
	RP3	I am willing to bear higher initial risks for green technologies with greater potential returns.
Environmental Strategic Orientation (ESO)	ESO1	The port regards green development as an important strategic objective.	Tseng et al. [118] Larabi [119]
	ESO2	Environmental protection occupies a central position in the port’s medium- and long-term strategy.
	ESO3	The port encourages the adoption of innovative technologies that contribute to emission reduction and environmental protection.
	ESO4	In decision making, the port tends to prioritize options with lower environmental impact.
Short-Term Performance Pressure (SPP)	SPP1	Due to short-term performance assessment requirements, it is difficult for us to allocate resources to green technology projects.	Mitchell et al. [120]
	SPP2	Short-term operational pressure constrains our ability to invest in carbon-reduction technologies.
	SPP3	Under short-term performance pressure, I tend to prioritize tasks with immediate results rather than green technology initiatives.
Technology Investment (TI)	TI1	We have a clear intention to increase investment in carbon-reduction technologies in the future.	Li et al. [44] Hu et al. [45]
	TI2	I am willing to promote greater resource allocation to green technologies within the port.
	TI3	We plan to gradually expand the application of green technologies over the next few years.
	TI4	If new green technologies emerge, we will actively consider their adoption.

). Among them, technology investment was operationalized using perceptual measures to capture port enterprise managers’ subjective investment intentions and strategic commitment toward carbon-reduction technologies. This approach was adopted for two main reasons. First, different types of carbon-reduction technologies vary substantially in investment scale, investment horizon, and accounting standards [5], making objective cross-firm comparisons difficult. Second, as this study focuses on managers’ decision cognition and resource allocation tendencies under uncertainty, perceptual investment intentions and proactive commitment are more direct indicators of underlying decision-making mechanisms than ex post financial expenditure outcomes.

5.2. Research Participants and Data Collection

This study collected data using a structured questionnaire, which consisted of four sections. Section 1 provided an overview of the study, briefly introducing the research objectives and content to respondents. Section 2 presented the informed consent statement, clearly informing participants of their rights, the intended use of the data, and data protection measures; respondents were allowed to proceed with the survey only after indicating their consent. Section 3 gathered demographic information, including gender, age, educational attainment, income level, and occupational category. Section 4 presented the measurement items designed around the core research variables. All items were measured using a five-point Likert scale (1 = strongly disagree, 5 = strongly agree). Compared with seven- or eleven-point scales, the five-point Likert format has been shown to effectively reduce respondents’ cognitive burden and minimize comprehension bias and method variance arising from excessive response options [84].

The selection criteria for survey participants were consistent with those applied in the preliminary qualitative interview phase. Following the recommendations of Hair et al. [85] regarding sample size requirements—namely, that each construct and its measurement items should be supported by 10–15 respondents—this study included nine latent constructs and 31 measurement items. Accordingly, the theoretically required effective sample size ranged from 310 to 465 respondents. After further accounting for an anticipated attrition rate of approximately 10% and the possibility of invalid responses, the targeted sample size was estimated to fall between 344 and 517 cases. In addition, a statistical power analysis was conducted using G*Power 3.1.9 (parameters: f² = 0.15, α = 0.05, Power = 0.95), which indicated a minimum required sample size of 138 respondents [86].

Data collection was conducted between 9 October and 28 November 2025. The research team adopted the same recruitment approach and participant eligibility criteria as those used in the qualitative phase, distributing questionnaires to 431 qualified practitioners. All participants were explicitly informed of the anonymity of the survey and the confidentiality of the data, and were reminded that there were no right or wrong answers. A total of 398 questionnaires were returned, with all respondents providing informed consent, yielding a response rate of 92.3%. After screening and excluding 26 questionnaires with substantial missing data or abnormal response patterns, 372 valid responses were retained for analysis. This sample size satisfies the minimum requirements proposed by Hair et al. [85] and Hair Jr et al. [86], and is adequate for subsequent statistical analyses.

5.3. Analytical Strategy and Control Variables

This study adopts a combined analytical approach integrating SEM and fsQCA. SEM generally comprises partial least squares SEM (PLS-SEM) and CB-SEM [87]. Given that CB-SEM places stronger emphasis on theory-driven model testing, allows for the simultaneous evaluation of measurement and structural models within an overall model-fit framework, and provides relatively high parameter estimation precision and robust significance testing for validating hypothesized paths as well as their direct and indirect effects [45], this study employs CB-SEM to conduct causal verification of linear net effects. Meanwhile, investment in carbon-reduction technologies by port enterprises is characterized by complex decision-making processes, in which the marginal effects of single variables are often insufficient to explain the empirical reality that “different combinations of resources can equivalently lead to high levels of investment,” reflecting strong equifinality and contextual dependence. Therefore, this study further introduces fsQCA to examine the joint explanatory power of condition configurations from a configurational perspective. Unlike SEM, which is grounded in symmetric linear relationships, fsQCA is based on set theory and Boolean algebra and emphasizes causal complexity, equifinal multiple pathways, and asymmetric logic [88]. It enables the identification of multiple sufficient configurations leading to high levels of technology investment, as well as differentiated configurations associated with low investment, thereby revealing the mechanisms underlying the principle of “multiple paths to the same outcome” and capturing boundary conditions and substitution effects [44]. Accordingly, CB-SEM provides rigorous tests of theoretical hypotheses and net effects, while fsQCA complements this by elucidating configurational drivers and causal asymmetry. Their integration, following a “validation–extension” logic, jointly enhances both the internal validity and contextual explanatory power of the study’s conclusions [89,90]. In addition, SPSS 26.0 was used to conduct descriptive statistics, scale reliability analysis, and Harman’s single-factor test. Amos 23.0 was employed to perform confirmatory factor analysis (CFA), while the PROCESS macro was utilized to further test the proposed mediation and moderation effects.

Furthermore, prior studies indicate that demographic characteristics such as age, gender, and work experience may influence individuals’ technology usage tendencies [87,91]. Accordingly, these variables are incorporated into the analytical model as control variables. Specifically, gender is coded as a binary variable (1 = male, 0 = female), while age and work experience are treated as non-interval categorical variables and handled using dummy coding. For each set of categorical variables, k − 1 dummy variables are retained, with one category designated as the reference group, in order to mitigate potential interference from individual-level extraneous differences in the estimation of core psychological path parameters.

6. Results

6.1. Sample Characteristics and Common Method Bias (CMB) Test

Table 2. Demographic Characteristics of the Survey Sample (n = 372).

Category	Item	Frequency	%
Gender Composition	Male	193	51.9
	Female	179	48.1
Age Distribution (years)	18–30	16	4.3
	31–40	133	35.8
	41–50	187	50.3
	Above 50	36	9.7
Educational Background	Associate Degree or Below	113	30.4
	Bachelor’s Degree	192	51.6
	Postgraduate Degree or Above	67	18.0
Functional Department Category	Operations Management	83	22.3
	Safety and Environmental Protection	61	16.4
	Strategy and Planning	74	19.9
	Investment and Financial Decision-Making	58	15.6
	Technology and Engineering Management	76	20.4
	Others	20	5.4
Work Experience (years)	1–5	8	2.2
	6–10	112	30.1
	11–20	131	35.2
	21–30	103	27.7
	Above 30	18	4.8

summarizes the demographic characteristics of the 372 valid respondents. In terms of gender composition, males accounted for a relatively higher proportion of the sample, with 193 respondents (51.88%). Regarding age distribution, the 41–50 age group constituted the largest segment, comprising 187 respondents (50.27%). In terms of educational attainment, a bachelor’s degree was the most common level of education, reported by 192 respondents (51.61%). With respect to departmental affiliation, positions related to technology and engineering management represented the largest share, with 76 respondents (20.43%). Finally, in terms of work experience, respondents with 11–20 years of tenure formed the dominant group, totaling 131 individuals (35.22%).

Given that the data in this study were collected entirely through self-reported questionnaires, CMB may theoretically be a concern [87]. To assess the potential impact of such bias, this study followed the procedure recommended by Podsakoff et al. [92] and employed Harman’s single-factor test. The unrotated principal component analysis yielded nine factors that jointly explained 73.983% of the total variance, with the first factor accounting for only 9.534% of the variance, which is well below the commonly accepted threshold of 50%. Based on these results, it can be preliminarily inferred that common method bias does not pose a serious threat to the validity of the data in this study.

6.2. Reliability and Validity Assessment of the Measurement Model

To examine the internal consistency of the measurement scales, this study used SPSS 27.0 to calculate Cronbach’s alpha coefficients for each construct. As shown in

Table 3. Reliability and Validity Analysis Results.

Variable	Item	Mean	SD	Factor Loadings	α	AVE	CR
PR	PR1	3.35	0.984	0.792	0.861	0.608	0.861
	PR2	3.34	0.975	0.751
	PR3	3.35	1.000	0.825
	PR4	3.33	0.940	0.749
MR	MR1	3.52	0.832	0.818	0.831	0.626	0.833
	MR2	3.52	0.870	0.838
	MR3	3.48	0.861	0.711
CR	CR1	3.32	0.995	0.786	0.856	0.599	0.856
	CR2	3.30	0.957	0.752
	CR3	3.27	0.966	0.808
	CR4	3.28	0.946	0.748
ER	ER1	3.32	1.003	0.822	0.835	0.634	0.838
	ER2	3.34	1.037	0.727
	ER3	3.30	0.960	0.836
SU	SU1	3.42	1.031	0.735	0.822	0.610	0.824
	SU2	3.39	0.999	0.820
	SU3	3.40	0.964	0.785
RP	RP1	3.59	0.952	0.767	0.842	0.639	0.841
	RP2	3.63	0.967	0.801
	RP3	3.60	0.927	0.829
ESO	ESO1	3.39	1.031	0.822	0.856	0.600	0.857
	ESO2	3.43	1.006	0.714
	ESO3	3.38	0.981	0.753
	ESO4	3.33	1.017	0.805
SPP	SPP1	3.59	1.006	0.867	0.871	0.693	0.871
	SPP2	3.56	1.043	0.806
	SPP3	3.55	1.038	0.824
TI	TI1	3.43	0.973	0.794	0.863	0.614	0.864
	TI2	3.39	0.989	0.745
	TI3	3.44	0.954	0.786
	TI4	3.38	0.932	0.808

, the alpha values of all latent variables ranged from 0.822 to 0.871, exceeding the recommended threshold of 0.70 [85], thereby indicating satisfactory internal consistency and reliability of the measurement scales. Subsequently, CFA was conducted using Amos 24.0. As reported in Table 3, all standardized factor loadings exceeded the commonly accepted criterion of 0.60, suggesting that the observed indicators adequately represent their corresponding latent constructs [85]. In addition, composite reliability (CR) values ranged from 0.824 to 0.871, surpassing the recommended cutoff of 0.70, while average variance extracted (AVE) values ranged from 0.599 to 0.693, all above the suggested threshold of 0.50. These results indicate that the measurement model exhibits strong convergent validity [93].

Furthermore, as shown in

Table 4. Discriminant Validity Test Results.

	TI	SPP	ESO	RP	SU	ER	CR	MR	PR
TI	0.784
SPP	0.173	0.832
ESO	0.218	0.141	0.775
RP	0.548	0.159	0.223	0.799
SU	0.594	0.148	0.114	0.441	0.781
ER	0.266	0.054	0.189	0.123	0.267	0.796
CR	0.412	0.077	0.200	0.349	0.517	0.235	0.774
MR	0.462	0.159	0.156	0.413	0.516	0.259	0.361	0.791
PR	0.465	0.062	0.167	0.352	0.525	0.209	0.556	0.401	0.780

Note. Bold values represent the square root of the AVE.

, the correlation coefficients between each construct and other constructs are all lower than the square root of the corresponding AVE values, providing additional evidence of satisfactory discriminant validity for the measurement model [93].

Following the evaluation criteria adopted by Zhou et al. [94], Ding et al. [87], and Hu et al. [45], the overall fit indices of the measurement model in this study all fall within acceptable ranges (see

Table 5. Model Fit Indices.

Indices	Recommended Threshold	CFA Mdel (9-Factor)	ULMF Test (9-Factor + Method Factor)	SEM Model
χ2/df	<3	1.317	1.236	1.734
GFI	>0.8	0.919	0.929	0.883
RMSEA	<0.08	0.029	0.025	0.044
IFI	>0.9	0.978	0.985	0.946
CFI	>0.9	0.977	0.985	0.945
TLI	>0.9	0.974	0.980	0.940

), thereby supporting the adequacy and robustness of the measurement framework. Building on this baseline, a common method factor was further introduced into the nine-factor CFA model, with all measurement items simultaneously loading onto their respective latent constructs as well as the common method factor [95], in order to conduct the unmeasured latent method factor (ULMF) test [96]. The results indicate that, after the inclusion of the common method factor, the overall model fit indices (see Table 5) did not exhibit substantial changes compared with those of the original model. This finding provides further evidence that common method bias does not constitute a serious concern in the data of this study [96].

6.3. SEM Path Analysis

The overall fit of the structural equation model met acceptable criteria (Table 5). The results of the path analysis (

Table 6. Direct Effects of Structural Paths.

Hypothesis	Path Direction	B	β	C.R.	p	Result
H1	PR → SU	0.365	0.396	6.593	***	Supported
H2	CR → SU	0.338	0.367	6.150	***	Supported
H3	SU → TI	0.451	0.467	7.721	***	Supported
H6	MR → RP	0.449	0.415	6.617	***	Supported
H7	ER → RP	0.038	0.043	0.752	0.452	Rejected
H8	RP → TI	0.391	0.411	7.095	***	Supported

*** p < 0.001, → means direct effect.

and Figure 2) indicate that personal resources (β = 0.396, p < 0.001) and conditional resources (β = 0.367, p < 0.001) exert significant positive effects on subjective utility. Material resources have a significant positive effect on risk preference (β = 0.415, p < 0.001), whereas the effect of energy resources on risk preference is not significant (β = 0.043, p = 0.452). Furthermore, both subjective utility (β = 0.467, p < 0.001) and risk preference (β = 0.411, p < 0.001) have significant positive effects on technology investment. Accordingly, Hypotheses H1–H3, H6, and H8 are supported, whereas Hypothesis H7 is not supported. Additional tests of the control variables indicate that age, gender, and work experience are not significantly associated with technology investment.

To examine the potential mediating roles of subjective utility and risk preference within the value cognition pathway and the strategic risk-taking pathway, this study employed the PROCESS macro in SPSS 27.0 and conducted exploratory mediation analyses using the bias-corrected percentile bootstrap method. Specifically, 5000 bootstrap resamples were generated, and the significance of indirect effects was evaluated based on 95% confidence intervals. An indirect effect is considered statistically significant when the confidence interval does not include zero [97]. As shown in

Table 7. Mediation Path Analysis Results.

Path	Effect Type	Effect	S.E.	95% CI		Type of Mediation
				LLCI	ULCI
A1: PR→SU→TI	Total Effect	0.3971	0.051	0.295	0.496	Partial Mediation
	Indirect Effect	0.1788	0.0316	0.1201	0.2447
	Direct Effect	0.2183	0.0542	0.1121	0.3236
A2: CR→SU→TI	Total Effect	0.3577	0.055	0.246	0.466	Partial Mediation
	Indirect Effect	0.1887	0.0320	0.1286	0.2522
	Direct Effect	0.1691	0.0576	0.0569	0.2843
A3: MR→RP→TI	Total Effect	0.4262	0.053	0.319	0.530	Partial Mediation
	Indirect Effect	0.1409	0.0310	0.0831	0.2055
	Direct Effect	0.2852	0.0554	0.1765	0.3935
A3: ER→RP→TI	Total Effect	0.2013	0.060	0.089	0.322	Reject
	Indirect Effect	0.0376	0.0275	−0.0144	0.0950
	Direct Effect	0.1638	0.0511	0.0655	0.2699

→ means direct effect.

, the bootstrap confidence intervals for Paths A1–A4 do not include zero. Among them, Paths A1–A3 exhibit a partial mediation structure, indicating that personal resources, conditional resources, and material resources retain significant direct effects on technology investment. In contrast, Path A4 reflects a full mediation structure, suggesting that energy resources influence technology investment exclusively through risk preference, with no significant direct effect observed.

This study further employed the PROCESS macro to examine the moderating effects of environmental strategic orientation and short-term performance pressure. The conditional effects analysis indicates that when environmental strategic orientation is at a low level (−1 SD), the positive effect of subjective utility on technology investment is significant (effect = 0.3168, p < 0.001, 95% CI [0.1951, 0.4386]), and the positive effect of risk preference on technology investment is also significant (effect = 0.2614, p < 0.001, 95% CI [0.1370, 0.3857]). When environmental strategic orientation is at a high level (+1 SD), the effect of subjective utility on technology investment is substantially strengthened (effect = 0.5653, p < 0.001, 95% CI [0.4631, 0.6675]), and the effect of risk preference on technology investment is likewise further amplified (effect = 0.6080, p < 0.001, 95% CI [0.4807, 0.7353]). Simple slope analyses (Figure 3) further demonstrate that the relationships between subjective utility, risk preference, and technology investment are stronger at higher levels of environmental strategic orientation, whereas these relationships are relatively weaker when environmental strategic orientation is low. Accordingly, Hypotheses H4 and H5 are supported.

In addition, the test of the moderating effect of short-term performance pressure reveals an opposite pattern. The conditional effects analysis indicates that when short-term performance pressure is at a low level (−1 SD), the positive effect of subjective utility on technology investment is significant (effect = 0.5691, p < 0.001, 95% CI [0.4666, 0.6715]), and the positive effect of risk preference on technology investment is likewise significant (effect = 0.5631, p < 0.001, 95% CI [0.4452, 0.6811]). However, when short-term performance pressure is at a high level (+1 SD), the effect of subjective utility on technology investment is significantly weakened (effect = 0.3238, p < 0.001, 95% CI [0.2036, 0.4439]), and the effect of risk preference on technology investment is also attenuated (effect = 0.3115, p < 0.001, 95% CI [0.1915, 0.4314]). Simple slope analyses (see Figure 4) further confirm that, compared with conditions of lower short-term performance pressure, higher levels of short-term performance pressure significantly weaken the positive relationship between subjective utility and technology investment, as well as the positive relationship between risk preference and technology investment. Accordingly, Hypotheses H9 and H10 are supported.

6.4. Configurational Analysis Based on fsQCA

The mediation analysis conducted using the PROCESS macro indicates that personal resources, conditional resources, material resources, and energy resources all exert significant positive direct effects on technology investment. However, because PROCESS is grounded in assumptions of linear relationships and average effects, it is limited in its ability to uncover potential nonlinear associations and interactive configurations among resource elements. A growing body of research suggests that technology adoption behavior is highly complex, and that individuals’ technology investment decisions are not driven by a single factor in a linear manner, but rather emerge from the joint effects of multiple conditions operating in different combinations [44,98]. Accordingly, this study further employs fsQCA to identify multiple equifinal resource configurations that lead to high levels of technology investment from a configurational perspective. This approach compensates for the limitations of traditional linear models and provides a more comprehensive representation of the complex mechanisms underlying technology investment behavior.

6.4.1. Data Set Calibration and Membership Assignment

Prior to conducting fsQCA, the data obtained using five-point Likert scales must be transformed into fuzzy-set membership scores ranging between 0 and 1 [87]. To this end, the study first calculated the mean values of the multiple measurement items for each latent construct to form composite indicators. Subsequently, following the calibration principles proposed by Ragin [99], the 5th percentile was set as the threshold for full non-membership, the 50th percentile as the crossover point, and the 95th percentile as the threshold for full membership. The membership scores were then computed using the Calibrate function in fsQCA 3.0. To avoid the exclusion of cases located exactly at the crossover point and to enhance the robustness of subsequent analyses, all membership scores equal to 0.50 were uniformly adjusted upward by 0.001 [100]. The final calibration results are presented in

Table 8. Fuzzy-Set Calibration and Descriptive Statistics.

Original Variables	Fuzzy-Set Calibration Thresholds			Calibrated Variables	Descriptive Statistics
	Full Membership	Crossover Point	Full Non-Membership		Mean	SD	Min	Max
PR	4.750	3.250	2.000	FPR	0.519	0.303	0.001	0.971
CR	4.667	3.667	2.333	FCR	0.388	0.300	0.001	0.981
MR	4.613	3.250	2.000	FMR	0.589	0.293	0.021	0.981
ER	4.667	3.333	1.667	FER	0.524	0.292	0.011	0.981
TI	4.750	3.500	2.000	FTI	0.492	0.305	0.011	0.971

.

6.4.2. Data Set Calibration and Membership Assignment

The necessity analysis aims to determine whether any predictor constitutes an indispensable condition for achieving a high level of technology investment. According to the criteria proposed by Dul [101], a predictor can be regarded as a “necessary condition” only when both its consistency and coverage exceed 0.90. As shown in the results presented in

Table 9. Results of the Necessary Condition Analysis for High Levels of TI.

Condition Variables	Consistency	Coverage
FPR	0.766	0.725
~FPR	0.556	0.569
FCR	0.613	0.777
~FCR	0.684	0.549
FMR	0.826	0.690
~FMR	0.491	0.587
FER	0.750	0.704
~FER	0.583	0.602

, none of the individual variables meets this threshold. Therefore, the attainment of high levels of technology investment does not depend on any single factor as a prerequisite.

6.4.3. Sufficiency Analysis

Following the analytical procedures proposed by Ragin [99], this study employed fsQCA 3.0 to construct a truth table comprising 2^k configurations, where k denotes the number of causal conditions. Each row of the truth table represents a possible configuration composed of seven condition variables and is accompanied by information on case frequency and consistency. Given that the sample size of this study exceeds 150, and in line with the recommendations of Fiss [100] and Pappas and Woodside [102], the frequency threshold was set at 3 and the consistency threshold at 0.85, in order to exclude configurations with insufficient empirical relevance or reliability. Based on these criteria, the fsQCA procedure generated three types of solutions: complex solutions, intermediate solutions, and parsimonious solutions. Because intermediate solutions retain strong explanatory power while providing a clearer representation of the underlying causal structure [99], this study adopts the intermediate solutions as the primary basis for interpretation. Subsequently, by comparing the intermediate solutions with their corresponding parsimonious solutions through counterfactual analysis, core and peripheral conditions within each configuration were further distinguished [100]. It should be noted that, in some cases, a single intermediate solution may correspond to multiple parsimonious solutions. In such instances, following the technical guidelines of Pappas and Woodside [102], conditions that repeatedly appeared in the parsimonious solutions were uniformly identified as core conditions.

Table 10. fsQCA Configurational Pathways Leading to TI.

Variables	Configurational Pathways
	S1	S2	S3
FPR			●
FCR	●	●
FMR		△
FER	△		●
Consistency	0.816	0.832	0.803
Raw Coverage	0.395	0.350	0.617
Unique Coverage	0.054	0.008	0.268
Solution Consistency	0.777
Solution Coverage	0.730

Note. ● indicates the presence of a core condition; △ indicates the absence of a peripheral condition; blank cells indicate that the condition is irrelevant or optional in the configuration.

summarizes three causal configurations that lead to high levels of technology investment. According to the criteria proposed by Dul [101], a configuration can be considered to have sufficient explanatory power when its consistency exceeds 0.80 and its raw coverage is no less than 0.20. All three configurations identified in this study satisfy these thresholds. The overall solution exhibits a consistency of 0.777 and a coverage of 0.730, indicating that, taken together, these configurations provide a stable and effective explanation of the mechanisms underlying high technology investment. Among them, Path S3 shows the highest raw coverage (0.617) and a relatively high consistency (0.803), suggesting that this configuration contributes substantially to explaining high technology investment. This path identifies high personal resources and high energy resources as core conditions, highlighting that strong individual capability and energy endowments constitute an important foundation for achieving high levels of technology investment. Path S2 exhibits the highest consistency (0.832) and a satisfactory level of raw coverage (0.350), indicating strong robustness and explanatory necessity. This configuration underscores the critical role of the combination of high conditional resources and low material resources in driving increased technology investment. Path S1 also demonstrates relatively high consistency (0.816) and raw coverage (0.395), suggesting that the synergistic effect of high conditional resources and low energy resources represents another key pathway toward achieving high technology investment.

7. Discussion

This study investigates the decision-making mechanisms underlying port enterprises’ investments in carbon-reduction technologies and yields systematic findings with strong explanatory power. First, with respect to RQ1, the results indicate that port enterprises’ carbon-reduction technology investment decisions involve multiple types of resources, including personal resources, conditional resources, material resources, and energy resources. This finding closely aligns with the resource classification framework proposed by COR theory [22], thereby confirming the applicability of COR theory in the context of port carbon-reduction technology investment.

Second, addressing RQ2, this study reveals a dual psychological mechanism through which resources influence technology investment, namely the “value cognition pathway” and the “strategic risk-taking pathway.” Within the value cognition pathway, personal resources significantly and positively predict subjective utility, which in turn promotes technology investment (supporting H1 and H3). This result is consistent with the findings of Chahal and Rani [54] and Zhao et al. [55]. From a COR perspective, personal resources, as a typical gain-type resource, enhances individuals’ perceived control over their environment and motivates them to pursue further resource gains [22]. This suggests that improving practitioners’ professional competence, cognitive capability, and psychological resilience can strengthen their perceived value of carbon-reduction technologies, thereby fostering investment behavior. Similarly, conditional resources indirectly enhance technology investment by positively influencing subjective utility (supporting H2), which is in line with Ebadi and Raygan [58] and Jeilani and Abubakar [59]. This finding indicates that policy, institutional, and informational support can reduce perceived uncertainty and threats of resource loss, thereby increasing managers’ value expectations. Such a mechanism is consistent with BDT, which posits that environmental conditions shape value trade-offs in decision-making [34]. Within the strategic risk-taking pathway, material resources significantly and positively predict risk preference, which in turn strengthens technology investment (supporting H6 and H8), consistent with Bagh et al. [68]. According to COR theory, abundant material resources expand decision-makers’ safety margins and reduce sensitivity to potential losses, thereby enhancing risk tolerance [20]. Combined with prospect theory, when expected returns outweigh perceived risks, managers are more inclined to engage in exploratory investment behaviors [33]. Accordingly, stable financial capacity and equipment availability increase port enterprises’ willingness to undertake strategic risk under technological uncertainty. However, energy resources do not significantly predict risk preference, a result that diverges from Molina-García et al. [73] and suggests the indirect and context-dependent nature of energy resources. First, as a capital-intensive industry characterized by strong regulation and institutional constraints [72], the port sector relies heavily on formalized approval procedures, budget constraints, and policy compliance in technology investment decisions. Under such conditions, even when managers possess sufficient time and effort to evaluate and coordinate projects, their risk attitudes may still be primarily constrained by material resource availability, accountability for investment failure, and organizational governance mechanisms. Second, in the context of port carbon-reduction investment, energy resources are more closely associated with operational and process support, mainly facilitating project implementation, interdepartmental coordination, and efficiency improvement, rather than directly shaping managers’ risk attitudes or orientations [103].

In addition, regarding moderating effects, environmental strategic orientation significantly strengthens the relationships among subjective utility, risk preference, and technology investment (supporting H4 and H5). This indicates that when organizations adopt a stronger green strategic orientation, expectations of resource gains become clearer and psychological mechanisms are more readily activated, which is consistent with Yin et al. [62] and Čater et al. [63]. In contrast, short-term performance pressure exhibits a weakening moderating effect on these relationships (supporting H9 and H10). According to the loss spiral mechanism of COR theory, high-pressure contexts intensify perceived resource depletion, prompting managers to prioritize risk avoidance and immediate returns, thereby constraining technological innovation and long-term investment [79,80]. The findings of this study provide empirical support for this theoretical logic.

Finally, with respect to RQ3, fsQCA identifies three equivalent configurational pathways leading to high levels of technology investment, complementing the results of linear analysis. The fsQCA findings demonstrate that energy resources are not “ineffective factors” under specific resource configurations; rather, they may function as complementary elements that, in combination with other resources, form viable pathways to high technology investment. This divergence does not indicate a contradiction between linear and configurational analyses but instead reflects different levels of theoretical explanation. Linear analysis identifies “necessary driving factors on average,” whereas fsQCA reveals “sufficient condition combinations under context dependence.” In certain configurations, individual resource elements may exhibit substitution effects, whereby deficiencies in one key resource can be compensated by combinations of other resources, thereby sustaining investment levels. For example, configuration S1 indicates that even with relatively low energy resources, a high level of conditional resources can still generate a strong technology investment tendency; similarly, S2 shows that even under low material resources, high conditional resources can trigger technology investment. Such substitution relationships are difficult to capture through linear models but are effectively revealed through configurational analysis. Therefore, this study concludes that port enterprises’ carbon-reduction technology investment does not follow a single optimal pathway but instead exhibits clear characteristics of causal complexity, equifinality, and multiple concurrent pathways.

7.1. Theoretical Implications

The theoretical contributions of this study can be summarized in three main aspects. First, this study introduces COR theory into the research context of carbon-reduction technology investment in port enterprises. It systematically identifies four categories of key resources underpinning such investment decisions—personal resources, conditional resources, material resources, and energy resources—and demonstrates that this resource structure is highly consistent with the resource classification framework proposed by COR theory [20]. By doing so, this study extends the application boundary of COR theory to long-term, strategic organizational decision-making and strengthens its explanatory power in the context of green transformation.

Second, this study advances theory by integrating BDT with COR theory and developing a dual-path analytical framework linking resources, psychological mechanisms, and technology investment. Specifically, it reveals how resource endowments influence port enterprises’ green technology investment through a value cognition pathway and a strategic risk-taking pathway. In contrast to prior studies that primarily adopt perspectives such as policy pressure [9,10], regulatory intensity [12,104], or economic and technical efficiency analysis [105,106], this study emphasizes managerial psychological mechanisms, highlighting the critical roles of bounded rationality and risk perception in green innovation decision-making. The findings not only confirm the mediating roles of subjective utility and risk preference in the relationship between resources and technology investment, but also demonstrate that environmental strategic orientation and short-term performance pressure exert significant moderating effects on the dual-path mechanism. In doing so, the study broadens the applicability of BDT within complex organizational settings.

Finally, from a methodological perspective, this study adopts a multi-method research design that integrates grounded theory with three-stage coding, CB-SEM, PROCESS analysis, and fsQCA, forming a comprehensive analytical paradigm of “qualitative identification–linear validation–configurational revelation.” This approach overcomes the limitations of traditional SEM, which tends to focus on average effects and single linear logic [44]. By adopting perspectives of causal asymmetry and equifinality, the study identifies multiple resource configuration pathways leading to high levels of technology investment, thereby uncovering the complexity and multi-path realization mechanisms underlying green technology investment behavior. This methodological contribution provides a valuable reference for future research on technology adoption, green innovation, and organizational strategic behavior.

7.2. Managerial Implications

This study provides important practical implications for port enterprises seeking to promote carbon-reduction technology investment under China’s “Dual-Carbon” strategy. First, the findings indicate that personal resources and conditional resources primarily influence investment decisions by enhancing managers’ subjective utility perceptions of green technologies, whereas material resources drive higher levels of technology investment by increasing risk preference. This suggests that, in the process of green technology adoption, strengthening managers’ capabilities in policy interpretation, technology evaluation, and strategic judgment, together with establishing institutionalized organizational support systems, constitutes a critical foundation for promoting technology investment [107]. Accordingly, when implementing the Dual-Carbon strategy, port enterprises should prioritize managerial capability development, the construction of supportive institutional frameworks, and the enhancement of information transparency. By reducing decision-making costs, strengthening perceptions of technological value, and combining these efforts with flexible capital allocation, asset upgrading, and budgetary slack, firms can effectively expand their investment safety margins.

Second, the study finds that environmental strategic orientation significantly strengthens the transformation of resources into technology investment, whereas short-term performance pressure suppresses this process, indicating that strategic systems and performance systems play critical moderating roles in green technology investment. Port enterprises should clearly articulate green development objectives at the strategic level and systematically integrate carbon-reduction technologies into long-term planning [108]. At the same time, performance management systems should place greater emphasis on long-term incentives while relaxing short-term profit constraints, and establish pilot-based tolerance and risk-mitigation mechanisms to encourage technological experimentation [109]. From a policy perspective, regulators may support multi-path green transformation across ports with heterogeneous resource endowments through differentiated policy instruments such as tiered subsidies, green finance tools, and inter-port coordination mechanisms, thereby promoting the overall synergistic development of green ports in China.

Finally, the fsQCA results demonstrate that high levels of technology investment are not driven by any single key resource, but rather emerge from the joint effects of multiple resource types configured in different combinations, with no necessary condition identified. This finding of equifinality challenges the linear assumption that “more resources necessarily lead to higher technology investment” [59] and highlights that optimizing resource configurations is more important than resource scale alone. Accordingly, port managers should avoid path dependence on single-resource advantages and instead develop differentiated resource allocation patterns aligned with their specific resource endowments and strategic positioning. By constructing technology investment pathways that are highly matched to organizational contexts, port enterprises can achieve sustainable and resilient green transformation.

7.3. Limitations and Future Research Directions

This study has several limitations. First, the sample is primarily drawn from coastal port enterprises in China, resulting in a relatively concentrated regional and industry context. As such, the external validity of the findings in different institutional environments or other infrastructure sectors remains to be further examined. Second, although the integration of CB-SEM and fsQCA enhances the overall explanatory power of the analysis, this mixed-methods framework is still based on static data structures and thus cannot fully capture the dynamic evolutionary processes among resource accumulation, psychological mechanisms, and investment behavior. Future research could address this limitation by adopting longitudinal or process-oriented research designs. Third, this study relies on managers’ self-reported perceptual measures of technology investment, which may be subject to cognitive bias or social desirability effects and do not directly reflect firms’ actual capital expenditures. Consequently, the findings primarily explain investment orientations and strategic commitments rather than concrete investment levels. Future studies could combine subjective assessments with objective financial data to enhance the robustness of the conclusions. Finally, this study treats carbon-reduction technology investment as an aggregate behavior and does not distinguish among heterogeneous technology types. Future research could further compare the differentiated decision-making mechanisms associated with digital energy management systems, shore power facilities, electrified equipment, and other low-carbon technologies, thereby improving analytical granularity and policy relevance.

8. Conclusions

Situated within the institutional context of China’s “dual-carbon” targets, this study integrates COR theory and BDT to develop a dual-path analytical framework—comprising a value cognition path and a strategic risk-taking path—that systematically elucidates the resource foundations, psychological mechanisms, and multi-path driving logic underlying carbon-reduction technology investment in port enterprises. The identified individual, conditional, material, and energy resources constitute a key resource portfolio for green investment in ports, providing empirical support for the applicability of COR theory in high-asset, high-uncertainty port contexts. Results from linear and conditional analyses indicate that individual and conditional resources promote technology investment by enhancing perceived utility, whereas material resources stimulate investment through increased risk preference; the direct effect of energy resources is relatively limited. Environmental strategic orientation and short-term performance pressure play reinforcing and inhibiting moderating roles, respectively. Further configurational analysis reveals that multiple resource combinations can equivalently drive high levels of investment, highlighting the causal complexity of green investment decision-making. Overall, this study deepens the understanding of port enterprises’ carbon-reduction technology investment decisions from a resource–psychological mechanism interaction perspective. Looking ahead to the research frontier of “sustainable maritime transport and logistics: efficiency, optimization, and decarbonization,” future studies may employ longitudinal data and multi-source objective indicators to capture the dynamic evolution of decision mechanisms, extend the analysis to coordinated decarbonization and efficiency optimization across port–shipping–logistics systems, and conduct comparative research across different institutional and market contexts to identify more generalizable and policy-actionable pathways for low-carbon transition.