Mitigating the Transaction Costs of Project Subcontracting Management: The Heterogeneous Effect of Behavior Control and Outcome Control

Abstract

Appropriate subcontracting strategy optimizes large-scale engineering project organization, enhancing implementation efficiency. However, numerous subcontractors increase the transaction costs of project subcontracting management. This study classifies subcontracting costs into coordination, monitoring, and negotiation costs based on transaction cost economics (TCE), and divides contractor management control into behavior and outcome control. This study aims to examine the causal mechanisms between subcontractor dispersion and the three types of transaction costs, plus the moderating effects of behavior and outcome control. Data is collected through a survey questionnaire and is analyzed by Partial Least Squares (PLS) analysis. Results show that the extent to which work is distributed among multiple subcontractors is positively related to transaction costs. Behavior control exercised by general contractors significantly suppresses the increase in transaction costs (especially the monitoring costs) resulting from subcontractor dispersion. However, outcome control does not exhibit a significant moderating effect on the relationship between subcontractor dispersion and transaction costs. The findings contribute to the theory of TCE and fill the gap in the research on the effective application of different management control modes. The findings help project managers choose an appropriate mode of management control to effectively manage subcontracting.

1. Introduction

This paper focuses on the ex-post transaction costs borne by the general contractor in delegated management contexts. As the complexity of the project increases, the number of subcontractors involved also grows. Subcontractor dispersion refers to the degree to which work is spread across multiple subcontractors [1], which is a critical component of subcontracting arrangements [2]. The appropriate degree of subcontractor dispersion can optimize the organization of engineering projects and then improve the implementation efficiency of the project, which is essential to achieve the project schedule, cost, and quality objectives. Implementing a higher dispersion strategy can promote competitive advantages in subcontracting market to obtain better contract prices [3,4] and may also reduce hold-up risks caused by excessive concentration [5,6]. However, due to the information asymmetry between general contractors and subcontractors, subcontractors often have opportunistic tendencies and behaviors, such as “rip-off” to maximize their interests, thus increasing the complexity of transaction activities, contractual risks, and transaction costs. Therefore, managing various subcontractors is challenging for general contractors.

According to transaction cost economics (TCE), management control helps to reduce transaction risk, increase the efficiency of inter-organizational cooperation, and has positive impact on transaction costs [7]. This concept comprises governance mechanisms and oversight processes designed to align partner’s conduct with predetermined project goals [2,8]. Thus, general contractors should implement effective management control in projects with several subcontractors to minimize transaction costs.

Management control provides tools to align inter-organizational interests and can be classified as behavior control and performance control [9,10]. Researchers evaluated different effects of control mechanisms separately and investigated the context under which they may function [2,11,12].

Currently, limited studies have focused on subcontractor dispersion and its impact on transaction costs, and relevant empirical analysis is lacking. Meanwhile, the relationship between subcontractor dispersion and transaction costs is uncertain under different management controls exerted by general contractors. It is noteworthy that in many project contexts, the owner often adopts a contractual model wherein they do not directly manage subcontractors but rather delegate their management to the general contractor. This study specifically focuses on the transaction costs and management strategies between the general contractor and its subcontractors. Although another approach might involve examining transaction costs between the owner and multiple subcontractors, obtaining sufficient samples directly from owners presents significant practical challenges. Therefore, this research concentrates on the relationship between general contractors and subcontractors.

In the following parts, based on the transaction cost theory, this study constructs a conceptual model to explore the impacts of subcontractor dispersion on transaction costs and how their relationships alter under different contexts of management control, thus filling the gaps in existing research. The findings contribute to the subcontracting management literature by jointly considering the effects of subcontractor dispersion and management control. This article will help general contractors by reducing transaction costs in subcontracting through the implementation of appropriate modes of management control.

2. Theoretical Background

2.1. Organizational Arrangement of Subcontracting

Subcontracting is an important strategic way for general contractors to undertake large-scale projects [13], which has economic rationality and is in line with the theory of comparative advantage of classical political economy [14]. Current literature primarily examines the selection and management of subcontractors. In terms of subcontractor selection, scholars identify selection criteria [15,16] and propose a variety of complex selection methods, including heuristic genetic algorithm, fuzzy set theory, hierarchical analysis, and data envelopment analysis [17,18].

In terms of subcontractor management, scholars explored how to coordinate the work among subcontractors and how to manage the relationship between the general contractor and subcontractor. These studies indicate that effective subcontractor coordination and relationship management can be achieved through relationships, contracts, and social governance [19,20,21], as well as by rationally distributing the interests of the general contractor and subcontractor [22]. Complex subcontractor selection and management work puts high expectations on general contractors, so some other researchers have looked at general contractor competence and management decision modeling. Yang et al. [23] identified the competencies of general contractors from the perspective of the entire life cycle value chain, including innovation, green construction, information application, communication and coordination, and project management, and revealed the structural relationship of these capabilities. Chen et al. [24] developed a multi-objective mixed integer linear programming model that optimizes supplier selection, inventory management methods, order quantities, and the option of dividing a material order as integrated decisions.

Some scholars have also identified the influencing factors of subcontracting decisions. Tarziján and Brahm [25] emphasized that the same company adopts different subcontracting strategies on different projects, influenced by the number of projects undertaken by the company, the complexity of the projects, and the company’s position in the local market. Ma et al. [26] identified outsourcing restrictions, strategic needs, and cost objectives as three key points to consider in outsourcing decisions. According to Drews [27], engineering projects are multifaceted organizations in which power and organizational identity limit the decision space for outsourcing, while the mutual constraints or synergies of firm capabilities and transaction efficiency factors ultimately determine outsourcing decisions.

In China, subcontracting relationships are governed primarily by the Contract Law of the People’s Republic of China and the Construction Law, which mandate written contracts, define liability boundaries, and regulate payment terms [28]. These legal provisions reduce ex-ante negotiation costs but may increase ex-post monitoring costs due to statutory compliance requirements. In the research field of construction management, a few researchers have already applied TCE in analyzing the subcontracting decision. González-Díaz et al. [29] found that firms tend to subcontract less as specificity grows, and firms tend to subcontract more as output heterogeneity and the use of intangible assets and capabilities increase. The influence of uncertainty, temporary lack of ability, and geographical dispersion is not significant. Bridge and Tisdell [30] proposed an integrated model based on TCE and resource dependence theory to explain the vertical boundary of construction enterprises, but this model has not been tested empirically.

Subcontractor dispersion refers to the spatial and organizational distribution of project tasks among multiple specialized firms, representing a critical dimension of modern subcontracting practices [3,6]. While transaction cost economics (TCE) has been applied to analyze subcontracting strategies in construction management, existing empirical research has primarily focused on the antecedents of subcontractor dispersion, such as the general contractor’s governance capabilities [3,4], project goals [4], and uncertainty in the external environment [3]. However, these studies have largely overlooked the consequences of varying dispersion patterns, particularly their implications for transaction cost dynamics. Therefore, it is necessary to conduct an empirical study to clarify the impact of subcontractor dispersion on transaction costs and its effective domains.

2.2. Transaction Costs of Subcontracting

According to Williamson [7], transaction costs occur when a service or product is transmitted across a separable technology interface. Transaction costs are inevitable in subcontracting [6]. Costantino and Pietroforte [31] state that for home building general contractors, the top three transaction costs with the highest probability of occurring during the subcontracting process are coordination costs (100%), inspection costs (70%), and monitoring costs (60%). Although these costs are difficult to calculate precisely, such intangible costs exist and have a significant impact on the investment efficiency of general contractors [6].

Using contract signing as a separating point, subcontracting transaction costs can be divided into two types: ex-ante transaction costs and ex-post transaction costs [32,33]. Ex-ante costs are procedural, consisting of information-seeking costs, targeting costs, and bargaining costs [34]. There is less opportunity for savings in this area for general contractors. In contrast, ex-post costs are unpredictable, including the expenses of monitoring performance and enforcing contractual terms, such as monitoring, coordination, conflict resolution, and negotiation [35]. Transaction-cost theory suggests that, under conditions of high asset specificity and uncertainty, ex-post governance costs tend to exceed ex-ante contracting costs [36], and ex-post costs are usually the general contractors’ primary concern [37,38]. Thus, this article concentrates on ex-post transaction costs and further investigates the relationship between subcontractor dispersion and ex-post transaction costs based on transaction cost theory.

Ex-post transaction costs can be primarily categorized into three categories: coordination, monitoring, and negotiation [2,39,40]. Coordination difficulties are related to external uncertainty and bounded rationality. Costs for coordination are thus incurred by communication and information exchange [40]. Coordination among participants is inevitable in the case of engineering and construction projects [37], such as clarifying mission requirements and delivery, optimizing resource allocation, handling variations, and adjusting the price index [37,41]. The inherent unpredictability of construction projects renders the coordination of multidisciplinary tasks particularly complex and demanding.

Monitoring costs regard transaction risk, which refers to the possibility that the party may seek its interests by damaging the other party’s interests through opportunistic behaviors. Information asymmetry, specific assets, and loss of control over resources increase transaction risk [40]. Monitoring and control are thus required to mitigate transaction risk by restraining opportunistic behaviors. Monitoring costs encompass expenditures and resources required to evaluate project deliverables, verify service quality among stakeholders, regulate team conduct, and optimally allocate resources to guarantee target achievement and contractual compliance [42].

Negotiation costs are related to inter-organizational conflict. Given the inherent complexity and unpredictable nature of construction projects, disputes between general contractors and subcontractors are inevitable [2]. At each phase of conflict resolution, negotiation plays a crucial role, with outcomes ranging from amicable settlements to complex and costly disputes [43]. Negotiation costs for conflicts encompass direct costs (e.g., cost of hiring lawyers, professional counseling, and litigation, liquidated damages) and indirect costs (e.g., extra time devoted to conflict resolution, damages to relationships, reputation, and trust) [2,38].

In the process of subcontracting, ex-post transaction cost is difficult to capture and receives less attention in the construction industry [44]. Nonetheless, with technological advancements and the creation of lean management systems, it is difficult to further optimize marginalized production costs; however, there is still a lot of opportunity for improvement in these invisible costs. Managers can recognize that the implications of subcontractor dispersion on a project are numerous by studying several dimensions of transaction costs, and they should evaluate all potential outcomes.

2.3. Management Control

As a fundamental component of the formal governance tools in the governance framework, controls play a key role in project success by reducing inter-organizational conflicts and discouraging opportunistic conduct [45]. Inter-organizational control research encompasses both wide and narrow viewpoints. From a narrow perspective, control is based on formal contractual restrictions on the cooperative behaviors of each party to a transaction, and terms like formal control, regulation, and contract are typically used to describe [46]. From a broad perspective, control also refers to the implications of cooperative culture, interpersonal relationships, and value orientation; terms like informal control, self-control, and social control are typically used to describe [8]. During the subcontracting process, the general contractor primarily focuses on formal control through contracts and rules [10,47]. Formal control consists of supervisory behavior and measurement of results [10]. Supervisory behavior, also known as behavior control or process control, is used to guide partners toward adopting behaviors that support goal achievement and to make sure they comply with the procedures, standards, and norms outlined in the contract [48]. Measuring results (outcome control, goal control) assesses how well the partner performed on the mission [49].

Within project management, two prevalent forms of managerial oversight are behavioral control and outcome-based control. Behavior control is a type of management control in which general contractors emphasize the importance of complying with rules and procedures by formulating behavior norms, work procedures, monitoring mechanisms, and evaluation criteria that expect the subcontractors to comply with [50]. The manifestation of behavior control includes (1) defining a set of construction rules and procedures; (2) describing task arrangements and responsibility sharing; (3) holding meetings and progress reports; and (4) inspecting and guiding the construction site. Different from behavior control, outcome control is a management control method oriented towards work results. The general contractor defines the expected goals, rewards, and penalties, focuses on the milestones and final results, and does not stipulate the methods and processes for subcontractors to complete the goals [50], which emphasizes more on the importance of completing the results [12]. It specifies the evaluation criteria for judging the subcontractors’ targets (e.g., stage gates, delivery schedules, and financial allocations) [2]. In contrast, under the outcome mode, subcontractors are evaluated based on how well they accomplish the goals [2,12,47,51]. General contractors can enhance the effectiveness of project management, foster cooperation and communication with subcontractors, and facilitate the seamless advancement of engineering projects by strategically enforcing behavior control and outcome control.

3. Development of Hypotheses

3.1. Transaction Cost Implications of Multi-Subcontractor Dispersion

The consequences of subcontractor dispersion will be explored from the perspectives of coordination, monitoring, and negotiation costs. As for the coordination costs, due to the limited rationality of humans, subcontractors are unable to obtain all project information, as well as fully understand the overall project objectives [37]. Thus, general contractors must dedicate substantial resources to synchronize subcontractor activities and ensure alignment with overarching project goals. When project tasks are allocated across multiple specialized subcontractors, the general contractor must play the role of a busy information center for communication, coordination, and decision-making because subcontractors are not able to access information regarding all aspects of the project. Furthermore, subcontractors are distinct organizations with different professional orientations, aims, and values, and the simultaneous involvement of many subcontractors will increase organizational interfaces and potential conflicts [52]. General contractors will expend a lot of time and energy to deal with these conflicts, which greatly increases the coordination strain. Conversely, if construction tasks are assigned to a small number of subcontractors, there are fewer organizational interfaces, which lowers the likelihood of conflicts. Additionally, inter-organizational coordination between general contractors and subcontractors will be internalized into intra-organizational coordination among the subcontractors, greatly alleviating the general contractor’s coordination efforts and reducing coordination costs. Building upon this theoretical foundation, we formulate the following research hypothesis:

H1a. 

Subcontractor dispersion will increase coordination costs.

Regarding the monitoring cost, with low subcontractor dispersion, the portion of work packaged to a single subcontractor increases, so the general contractor will not have control over the specifics of the work in the package and will have clear and concise criteria to evaluate the outcome. If the general contractors distribute the project to more subcontractors, they must make a great effort to set a verifiable standard for each subcontractor. The outcomes of these breakdowns are usually more difficult to verify because their functions rely on those of the other parts. Meanwhile, as the interfaces between subcontractors increase, opportunistic behavior will easily occur in the interfaces because the rights and obligations are usually obscure on these interfaces [53]. To mitigate opportunistic conduct, general contractors must enhance their oversight mechanisms. Moreover, directly managing numerous subcontractors may be beyond the management span of the general contractor; thus, control and supervision may become inefficient, resulting in higher monitoring costs. Therefore, the following hypothesis is proposed:

H1b. 

Subcontractor dispersion will increase monitoring costs.

Since different parties in a construction project possess varying needs, goals, or interests, conflicts of interest and disputes between general contractors and subcontractors are inevitable during transactions [43]. Addressing these challenges necessitates expenditure on negotiation processes, encompassing both tangible expenses (such as legal and consulting fees) and intangible costs (including operational inefficiencies, schedule disruptions, and interpersonal conflicts among stakeholders) [2]. High subcontractor dispersion indicates more participants in the project, which increases the likelihood of conflicts of interest and points of dispute between general contractors and subcontractors compared to the case of low subcontractor dispersion. To address these problems, both parties must be adequately prepared before the negotiation and conduct a significant amount of bargaining work (changes, claims, etc.) for the precise phrasing and interpretations in the contract, which increases the difficulty of the negotiation and the cost spent. Based on this, we constructed the following hypothesis:

H1c. 

Subcontractor dispersion will increase the negotiation costs.

3.2. Moderation Effect of Behavior Control

As a typical temporary organization, engineering projects require members of different professional backgrounds and skills to work together to accomplish tasks. Behavior control can effectively improve this situation by regulating the performance behavior of the participants [54]. General contractors clarify work specifications, procedures, task scheduling, and responsibility allocation, and provide behavior guidance to subcontractors to ensure that they complete the work tasks according to the expected requirements [50]. This lessens the time and effort the general contractor must devote to communication and coordination, as well as reduces conflicts and negotiation costs due to miscommunication or non-compliance with work processes. Simultaneously, the general contractors obtain behavior information of subcontractors through construction sites, direct monitoring, and reporting [51], and timely require subcontractors to revise assignments that do not achieve the expected outcomes, to avoid additional transaction costs resulting from subcontractor errors or speculative behavior. Furthermore, subcontractors’ rewards are contingent on their behavior and work completion. This incentive mechanism encourages subcontractors to comply with work processes and standardize their work behaviors [55], reducing the general contractor’s effort in monitoring and negotiation. Therefore, we proposed the following hypothesis:

H2a. 

Behavior control weakens the influence of subcontractor dispersion on coordination expenditures.

H2b. 

Behavior control weakens the influence of subcontractor dispersion on monitoring expenditures.

H2c. 

Behavior control weakens the influence of subcontractor dispersion on negotiation expenditures.

3.3. Moderation Effect of Outcome Control

Outcome-based controls define target performance metrics while granting subcontractors autonomy in determining implementation methods [2,12]. Thus, subcontractors can select suitable techniques and procedures to complete their job assignments throughout project implementation, depending on their professional expertise and experience. This minimizes the requirement for general contractor coordination, consequently cutting down on coordination expenses. As a result, the general contractors can focus more on project management, eliminating the requirement for monitoring subcontractors and thereby lowering monitoring expenses. In the meantime, detailed and precise standards for output performance could limit the uncertainty surrounding the responsibilities and functions of each party [2,47,56]. Moreover, it demonstrates trust in the subcontractor’s integrity, dependability, and competence [51], which may foster trust between the parties, incentivize subcontractors to work more conscientiously, avoid opportunistic tendencies, and reduce incidents of conflict, all of which can lower ex-post transaction costs. Thus, the hypothesis is as follows:

H3a. 

Outcome control weakens the influence of subcontractor dispersion on coordination expenditures.

H3b. 

Outcome control weakens the influence of subcontractor dispersion on monitoring expenditures.

H3c. 

Outcome control weakens the influence of subcontractor dispersion on negotiation expenditures.

The conceptual framework for the study is shown in Figure 1.

4. Method

4.1. Sample Identification and Data Collection

This study, to explain the relationship between subcontractor dispersion and transaction costs, is an empirical study relying on a large sample. Currently, no existing datasets systematically compile construction project data pertinent to our investigation’s focus. To ensure an adequate sample size, this study employed digital surveys administered to Chinese construction practitioners attending four university-organized project management training sessions. The study targets subcontractors within the Chinese construction industry, specifically those engaged in civil engineering, architectural projects, and specialized fields such as electrical and mechanical installation. These firms typically operate under fixed-price contracts and are subject to the regulatory framework of the Chinese construction market. The questionnaire was developed through a thorough review of past literature and semi-structured interviews with industry practitioners. Then the questionnaire was tested on three pilot projects to verify that respondents could correctly interpret the questions being asked, which resulted in several changes to the wording of specific questions to fit the construction project better.

Obtaining a genuinely random sample of construction projects proved impractical due to the inherent challenges in defining a complete sampling population for this unit of analysis. Thereby, this study employed a large-scale convenience sample in line with plenty of extant research [57,58]. All selected participants were industry veterans representing companies included in Engineering News-Record’s (ENR) prestigious Top 250 International Contractors list. Selected from the general contractors, the target respondents were professionals who were involved in managing subcontractors and had access to the expense and cost information of the project. Participants were instructed to identify a recently finalized project where their companies served as general contractors, and they were extensively engaged as well. The survey instrument explicitly stated that responses were not subject to correctness evaluation and guaranteed complete anonymity and data confidentiality. The questionnaire started with basic information about the project and respondents. Project distribution in the sample covered five primary types (31.6% for liner engineering, 13.4% for housing, 16.3% for port and waterway, 29.8% for energy development, 8.7% for industrial engineering), and size (5.4% for less than 30 million RMB, 12% for 30–100 million RMB, 57% for 100 million–1 billion RMB, 10.5% for 1–3 billion RMB, 14.9% for more than 3 billion RMB).

In total, among the 350 distributed questionnaires, 339 were collected with a response rate of 96.9%. Considering that the data were collected from training programs, the high rate of response is reasonable. Responses from unqualified participants, including technicians and other personnel with fewer than three years of professional experience, were excluded from analysis due to potential limitations in their technical expertise to provide valid assessments. The final dataset consisted of 275 usable responses, yielding a response validity rate of 81.12%. The sample characteristics are shown in

Table 1. Characteristics of respondents and projects.

Characteristic	Category	Frequency	Percentage (%)
Project type	Liner engineering	87	31.6
	Housing	37	13.4
	Port and waterway	45	16.3
	Energy development	82	29.8
	Industrial engineering	24	8.7
Work experience (years)	<3	29	10.5
	3–5	47	17
	6–8	56	20.3
	9–11	41	14.9
	>11	102	37
Size	Less than 30 million RMB	15	5.4
	30–100 million RMB	33	12
	100 million–1 billion RMB	157	57
	1–3 billion RMB	29	10.5
	More than 3 billion RMB	41	14.9
Job position	Project manager	34	12.3
	Contract and business manager	143	52
	Technician	63	22.9
	Others	35	12.7

.

4.2. Measures

The measurement of subcontractor dispersion was operationalized through three key dimensions: the quantity of subcontractors, contract packages, and work interfaces, utilizing the framework established by Hui et al. [1]. Whereas subcontractor dispersion may be naturally disparate across different types of projects, respondents were required to make a comparison with similar projects they had participated in. As for transaction costs, it is difficult to measure them precisely. However, there are a few measurement instruments to reflect them. The measurement of coordination costs was operationalized using a four-item instrument adapted from Clemons et al. [39]. Monitoring costs were assessed using a four-item scale adapted from Chen et al. [42]. Negotiation costs were measured by a five-item scale from Dahlstrom and Nygaard [59] and Mesquita and Brush [60]. The assessment of behavioral control utilized a four-item instrument adapted from Stouthuysen et al. [12], and outcome control was based on the scales from Heide et al. [11]. Survey instruments were contextualized for the construction sector by adopting the conceptual framework and nomenclature proposed by Ning [61].

To refine the survey items, we conducted a pilot study through interviews with three industry experts and one academic specializing in project contract governance. The measurement items were revised according to expert feedback to enhance clarity and practical relevance. Utilizing a seven-point Likert scale (1 = strongly disagree to 7 = strongly agree), these refined instruments are detailed in

Table 2. Measurement scales and their evaluation.

Construct and Measure Items	SFL
Subcontractor dispersion (α = 0.811; AVE = 0.724; CR = 0.887)
In contrast to comparable projects, this project involves a large number of subcontractors with contractual relationships with us.	0.862
2. In contrast to comparable projects, we segment this project into managing contract packages.	0.860
3. In contrast to comparable projects, there are many work interfaces between participants.	0.831
Coordination cost (α = 0.813; AVE = 0.650; CR = 0.881)
Throughout the project execution phase, significant resources were allocated to stakeholder engagement and knowledge dissemination among all participating entities.	0.827
2. Throughout the project execution phase, proactive measures were instituted to ensure timely reporting of operational data and real-time modifications by all contracted parties.	0.848
3. Throughout the project execution phase, to ensure all contractors understood project requirements, regular meetings and communications were maintained with each specialty firm.	0.837
4. Throughout the project execution phase, frequent discrepancies, conflicts, and disagreements emerge between stakeholders, requiring active mediation by the project team.	0.705
Monitoring cost (α = 0.849; AVE = 0.685; CR = 0.896)
Serious concerns exist about subcontractors potentially manipulating or concealing information for strategic advantage.	0.717
2. Significant effort is required to verify subcontractor compliance with specifications and quality standards.	0.837
3. Substantial resources have been allocated to project performance monitoring and control.	0.893
4. Significant resources have been allocated to monitor and regulate subcontractor conduct.	0.853
Negotiation cost (α = 0.864; AVE = 0.643; CR = 0.899)
Frequent and often severe conflicts arise between our organization and contracted partners, with many disputes escalating to significant levels of contention.	0.823
2. To effectively manage contractual disputes, our team consistently engages in thorough preparatory work prior to negotiation sessions to facilitate decisive outcomes.	0.631
3. Multiple intensive negotiation rounds were conducted to mediate contractual disagreements before ultimately achieving resolution.	0.824
4. Conflict resolution processes frequently involve rigorous negotiations concerning the precise phrasing and interpretation of contractual requirements.	0.863
5. Conflict resolution negotiations are frequently characterized by high levels of psychological stress and tension among participating parties.	0.845
Behavior control (α = 0.796; AVE = 0.624; CR = 0.869)
During the construction phase, detailed operational protocols were established and mandated for strict compliance by all contracted parties.	0.796
2. During the construction phase, contractor work methods were systematically revised when actual outcomes failed to meet projected performance targets.	0.688
3. During the construction phase, performance feedback and operational data were systematically communicated to contractors to facilitate necessary process improvements.	0.853
4. During the construction phase, rigorous monitoring was conducted on contractors’ execution strategies and work schedules to ensure compliance with project requirements.	0.814
Outcome control (α = 0.808; AVE = 0.648; CR = 0.880)
During the construction phase, well-defined performance metrics were established to guide contractors’ operational execution.	0.840
2. During the construction phase, systematic tracking was conducted to evaluate contractors’ performance against established targets.	0.831
3. During the construction phase, comprehensive milestone schedules were established to serve as monitoring benchmarks.	0.839
4. During the construction phase, a performance-based incentive system was implemented, tying contractor compensation directly to the attainment of predefined objectives.	0.702

Notes: SFL = standardized factor loading, α = Cronbach’s alpha, AVE = average variance extracted, CR = composite reliability.

.

4.3. Analytical Method

In this research, Partial Least Squares (PLS) is selected to analyze the data and test the validity of the hypothesis and model. At present, the PLS method is widely used in the field of structural equation modeling, which has low requirements on sample size and residual, and is suitable for the detection of small sample models [62]. Data are collected and arranged using a questionnaire survey, and SPSS 22.0 software is used for reliability and validity tests. A structural equation model is established by Smart PLS 4.0 software, and relevant data is calculated by the PLS Algorithm and Bootstrapping to analyze the path coefficient and validity of the model.

4.4. Construct Validity and Reliability

The reliability of the questionnaire is demonstrated through internal consistency testing, which assesses the inter-item correlation among measurement indicators. Through SPSS 22.0 software, the KMO value is 0.883, greater than 0.7. Bartlett’s sphericity test results show that the approximate chi-square is 3482.415, the degree of freedom is 276, and the Sig value of the significance probability of the sample data is 0.000, which is less than 0.05, confirming the correlation of the sample data.

Table 2 displays the measurement item specifications along with their reliability and validity test results. The reliability analysis yielded Cronbach’s alpha coefficients above 0.7 for every measured construct. The measurement model exhibited strong psychometric properties. All indicator loadings exceeded the 0.5 threshold, while composite reliability (CR) scores surpassed the 0.7 criterion for every construct. Furthermore, average variance extracted (AVE) values met the 0.5 benchmark, confirming adequate convergent validity. Discriminant validity was tested by comparing the square root of AVE with the off-diagonal correlation coefficients [63]. As shown in

Table 3. Means, standard deviations, and intercorrelations.

Constructs	Means	SD	Construct Correlations
			SD	CC	MC	NC	BC	OC
SD	3.948	1.719	0.851
CC	5.236	1.213	0.314 **	0.806
MC	4.900	1.326	0.187 **	0.637 **	0.828
NC	4.342	1.340	0.204 **	0.532 **	0.582 **	0.802
BC	5.476	1.186	0.051	0.381 **	0.231 **	0.089	0.790
OC	5.637	1.150	0.009	0.371 **	0.158 **	0.065	0.701 **	0.805

Notes: The values in boldface signify that they are greater than the off-diagonal correlations. * p < 0.05; ** p < 0.01, *** p < 0.001. The lower triangle is the correlation coefficient, and the diagonal is the square root value of AVE. SD = subcontractor dispersion; CC = coordination cost; MC = monitoring cost; NC = negotiation cost; BC = behavior control; OC = outcome control. The same is below.

, the square root value of AVE for each construct was greater than the off-diagonal correlation coefficient, providing evidence of discriminant validity.

5. Results

5.1. Structural Model Validity Testing

The predictive validity of the structural equation model was assessed using the blindfolding procedure in Smart PLS 4.0. Q2 of endogenous latent variables are 0.420, 0.474, 0.450, 0.377, and 0.412, respectively. The GOF value calculated by the formula is 0.33, which is close to 0.36. Refer to the study of [62], which meets the requirements. Therefore, the structural equation model in this study has passed the validity test.

5.2. Hypothesis Analysis

The regression analyses were conducted in two steps. Firstly, the main effects were tested.

Table 4. Main effects test results.

Hypothesis	Path	Original Sample	t-Value	p-Value	Results
H1a	SD→CC	0.294 ***	5.398	0.000	S
H1b	SD→MC	0.185 **	3.003	0.003	S
H1c	SD→NC	0.228 *	2.325	0.020	S

Note(s): S = supported, NS = not supported. The same is below.

summarizes the results. The bootstrapping algorithm in Smart PLS 4.0 software was used to calculate path coefficients for 5000 repeated samples and test the significance. A significantly positive relationship is found between subcontractor dispersion and coordination costs (β = 0.294, p = 0.000), monitoring costs (β = 0.185, p = 0.003), and negotiation costs (β = 0.228, p = 0.020), which supports H1a, H1b, and H1c. To sum up, the higher the dispersion of subcontractors, the higher the transaction cost of subcontracting.

Secondly, following Hayes’ [64] methodological framework, interaction terms were incorporated to analyze the moderating influences of behavioral control and outcome-based control mechanisms. This study examines how behavior and outcome controls moderate the relationship between subcontractor dispersion and three types of transaction costs, employing Bootstrapping operations for analysis. As shown in

Table 5. Moderating effects test results.

Hypothesis	Path	Original Sample	t-Value	p-Value	Results
	SD × BC→TC	−0.172 *	2.012	0.044
	SD × OC→TC	0.022	0.296	0.767
H2a	SD × BC→CC	−0.060	0.676	0.499	NS
H2b	SD × BC→MC	−0.179 *	1.986	0.047	S
H2c	SD × BC→NC	−0.173	1.625	0.104	NS
H3a	SD × OC→CC	0.018	0.203	0.839	NS
H3b	SD × OC→MC	0.036	0.418	0.676	NS
H3c	SD × OC→NC	0.010	0.136	0.892	NS

, general contractors’ behavior control of subcontractors plays a negative moderating effect on the overall effect of subcontractor dispersion on transaction costs (β = −0.172, p = 0.044), while the moderating effect of outcome control in the overall effect of subcontractor dispersion on transaction costs is not significant (β = 0.022, p = 0.767). Behavior control has a negative moderating effect on the effect of subcontractor dispersion on monitoring costs (β = −0.179, p = 0.047), in support of H2b. The moderating effects of the other paths are not significant, which does not provide support for H2a, H2c, H3a, H3b, and H3c. To further clarify and understand the moderating effect of behavior control, this study uses simple slope analysis to draw a diagram of the moderating effect, as shown in Figure 2. Behavior control was divided into two groups according to one standard deviation above and below the average score, and the effects of subcontractor dispersion on monitoring cost and transaction cost under different behavior control levels were analyzed according to the regression equation. Panel (a) shows the significant moderating effect of behavior control between subcontractor dispersion and monitoring costs, and panel (b) shows the significant moderating effect of behavior controls between subcontractor dispersion and transaction costs.

6. Discussion

Based on the above analysis, the theoretical model has been revised accordingly. The updated model, as illustrated in the figure, retains only the paths that were found to be statistically significant, in order to better reflect the actual influence mechanisms among the variables. Figure 3 is presented below for further clarification.

6.1. The Direct Effect

The results show that subcontractor dispersion and coordination costs are positively correlated. It is consistent with the existing research that a great degree of outsourcing will increase communication and coordination difficulties [6,39,65]. The relationship between subcontractor dispersion and monitoring costs is also positively related, in line with existing research [6,53]. The results of higher transaction costs also respond to the organization’s inefficiency in service outsourcing deconcentration found by Magelssen et al. [66]. A great degree of outsourcing also increases conflict negotiation costs. As the number of subcontractors increases, bargaining and disputing are more likely to occur and become more complicated, increasing the expenditure of ex-post transaction costs.

6.2. The Moderating Effect

The analysis reveals that behavioral governance significantly attenuates the positive relationship between subcontractor geographic distribution and transaction costs, with particularly pronounced effects on monitoring costs. In contrast, outcome control demonstrates no statistically meaningful moderating influence. This insignificance may stem from the high interdependence among subcontractors’ outcomes, which makes it difficult to attribute performance credibly to individual entities, thereby diluting the efficacy of outcome-based mechanisms. Furthermore, the ineffectiveness of outcome control aligns with the findings of Stouthuysen et al. [12], which demonstrated that the efficacy of control mechanisms is contingent upon the nature of the service and task environment. In contexts characterized by high dispersion and interdependence, outcome controls often prove ineffective due to significant contextual and cultural differences. These controls typically rely on standardized metrics that do not adequately account for varied local practices, communication norms, and relational expectations, thereby reducing their moderating utility. These conclusions are consistent with previous research [67] proposed that behavior and outcome controls would increase program performance. They discovered that behavior control had a direct effect on performance, while outcome control had no direct effect on performance. A plausible explanation is that in contexts involving dispersed subcontractors, outcome measures become noisy and conflated with external influences, reducing their value as control instruments. In the project management practice, general contractors frequently pay more attention to outcome control than behavior control. Nevertheless, the empirical analysis results indicate that the general contractors’ outcome control over subcontractors is ineffective in decreasing transaction costs when there is a high dispersion of subcontractors. Consequently, if more subcontractors are employed to complete the project, the general contractor should pay more attention to the behavior of subcontractors, ensure they follow standard operating procedures and norms, clarify their respective tasks and responsibilities, and improve communication and coordination. Concurrently, general contractors must maintain ongoing oversight of subcontractor performance to verify compliance with stipulated standards, thereby minimizing transactional expenditures.

7. Conclusions and Implications

This study shows that in large-scale engineering projects, having more subcontractors spread out leads to higher costs for coordination, monitoring, and negotiation. However, closely managing how these subcontractors work can help control these increased costs—especially by monitoring expenses. On the other hand, simply focusing on outcomes does little to reduce these costs.

These findings indicate that extensive reliance on outcome-based performance metrics is insufficient for managing transaction costs under the situation of high subcontractor dispersion. Instead, project managers should prioritize process-oriented behavior control. This entails actively monitoring subcontractor conduct, ensuring adherence to standard procedures, clarifying responsibilities, and enhancing coordination to reduce transaction costs effectively. However, the findings should be interpreted with caution when generalizing to other cultural or regulatory environments, such as those in Europe or North America, where contract enforcement mechanisms, market maturity, and subcontracting practices may differ significantly [68].

7.1. Theoretical Implications

Previous studies on project subcontracting mainly focused on the make-or-buy decision, but were seldom concerned with the subcontractor dispersion [1]. While prior research emphasized the antecedents of dispersion, this study examines its consequences on transaction costs, particularly in complex projects with high spatial-organizational distribution, filling a critical gap in TCE-based subcontracting governance. This research is a preliminary investigation of the consequences of subcontractor dispersion; it can be unified with previous studies on make-or-buy decisions and provides a basis for drafting a clearer picture of multi-step subcontracting strategies in construction projects. Among the various theoretical perspectives, this research adopts TCE to analyze the project-level subcontractor dispersion; thus, transaction costs are a major concern, which expands the application of TCE in the research field of construction management.

Meanwhile, this study finds that in scenarios where subcontracting is adopted, behavioral control is a more appropriate management control mode. Identifying the applicable scenarios for different management control models fills the gap in the research on the effective domain of management control modes.

7.2. Managerial Implications

In case of complex projects, if the general contractor only focuses on outcome control of various subcontractors, the subcontractors may fail to comply with the required regulations to meet the project goals, increasing subsequent transaction costs. To address the negative impacts of project subcontracting, a suitable behavior control mechanism may be implemented to generate a win-win scenario.

Digital management has progressively developed into a new trend of project management, as well as an effective way of controlling the behavior of subcontractors. Through the construction of a digital management platform, information exchange and communication can be increased, and transaction risk caused by information asymmetry can be mitigated. Thus, a digital management platform could be used as an important tool of behavior control to decrease the transaction costs.

7.3. Limitations and Directions for Future Research

This research still has some noteworthy limitations. Firstly, this research only considered transaction costs, but various objectives should be weighed when selecting the subcontracting strategy in a real decision-making situation. For instance, considering production efficiency, subcontracting activities to more professional organizations will improve the production efficiency of projects. In future studies, multiple objectives could be considered when examining the subcontracting dispersion. Second, the primary moderating factors used in this study are the two management control methods. Nonetheless, the contractual form and general subcontractors’ cooperative relationship will affect the subcontracting transaction cost; it must be further explained and investigated in the future to construct a unified framework for subcontracting construction projects. Third, measuring subcontractor dispersion uses subjective comparisons to “comparable projects,” which may cause bias. Future research should develop metrics based on management difficulty—such as the number, location, or trade diversity of subcontractors—rather than financial values alone. Additionally, this study did not account for subcontractors’ historical performance records, which may influence transaction costs—as more reliable subcontractors typically entail lower ex-post costs. However, such data were not accessible through our survey method; future studies could incorporate dedicated performance databases to control for this factor. Fourth, the convenience sampling of Chinese contractors may limit generalizability due to state-led governance, unique relational practices (e.g., guanxi), and distinct bidding mechanisms [69]. These context-specific factors constrain direct applicability to other institutional settings (e.g., Europe or North America) [70]. Future research should validate our findings in diverse cultural and regulatory environments to enhance their external validity.