Adaptability in Public Procurement of Engineering Services Promoting Carbon Reduction: An Organizational Control Perspective

Abstract

Many studies have emphasized the importance of engineering services and their adaptability to reduce carbon emissions in the construction sector. As public clients procure these services, selecting efficient (procurement-related) control modes is critical. However, studies of control modes and their effect on adaptability are scarce. The purpose of this study is to investigate how, through the choice of control modes, a public client may create incentives for service providers to deliver carbon reduction solutions in the construction phase. In this study, a procurement model and a classification model are developed, and a single case study of an engineering service contract aiming for carbon reduction is used to illustrate the models. The empirical data include 16 interviews with respondents from the 2 contract parties, i.e., the client and service provider. The findings show that it is important to create incentives for adaptability when procuring engineering services with the aim of reducing carbon emissions in the construction phase. The findings also highlight the importance of these incentives being created both in relation to the selection of service providers and to the responsibility between the contract parties. The models developed in the study may serve as important tools for both practitioners and researchers.

1. Introduction

Alarming reports from the Intergovernmental Panel on Climate Change (IPCC) have incentivized policymakers worldwide to reduce the increasing levels of greenhouse gases (primarily carbon) in the atmosphere. In the Paris Agreement, it is stated that both private and public actors have a vital role to play in reaching these targets [1]. Prior research shows that public procurement can serve as an important driver for promoting sustainability and reducing climate impact, which is sometimes referred to as green public procurement [2,3,4].

Increasingly, the infrastructure sector is seen as a major source of carbon emissions [5,6], and thus it has an important role to play in meeting national and international reduction targets [7]. In fact, the construction, maintenance, and operation of buildings and infrastructure account for 38% of carbon emissions globally [8]. A large share of these carbon emissions come from the transport of soil masses and construction materials and from maintenance and site operations [6].

Since engineering consultants are usually involved in shaping future infrastructure, planning [9], and design [10], their services are of great importance in solving sustainability issues such as reducing carbon emissions. Engineering services are usually provided by private engineering consultancy companies (ECCs) [10]. Due to the fact that clients procure these services, the role of clients in general [11,12,13], and their procurement strategies in particular, are of great importance in meeting carbon reduction targets [13]. However, it can be difficult for clients to choose procurement strategies (ibid) and monitor their effects [14].

Organizational control is widely designed and applied in inter-organizational relationships to influence and regulate the cooperating partner, coordinate and solve contractual issues, and achieve organizational objectives [15,16]. The control framework by Ouchi [17], which includes three control modes, i.e., market, bureaucracy, and clan (also referred to as output, process, and social control, respectively), is the most widely used within organizational control research [18,19]. Eriksson [20] and Eriksson and Laan [21] developed a procurement model for procuring contractors based on Ouchi’s control framework [17], which includes different procurement stages and different alternatives linked to the three control modes within each stage. A client’s procurement strategy therefore consists of a combination of (procurement-related) control modes [13,20,21]. Due to the relevance and use of organizational control in the research on inter-organizational relationships, including issues of procurement [15,16,22,23], the three control modes constitute the main theoretical perspective on the focal point of this study, i.e., the public procurement of engineering services to promote carbon reduction.

In general, the design process is considered uncertain, iterative [24,25], and organic [11], which makes adaptability desirable throughout the process [10,11,25,26]. Since carbon reduction is uncertain and complex, innovation is needed [27], which creates a need for adaptability to enable innovative solutions [11,26,28]. In addition to adaptability, it is important for public clients to provide incentives for service providers to deliver innovative solutions for carbon reduction [26]. However, clients need more knowledge about how to properly incentivize service providers [29]. In fact, a literature review on green public procurement shows that specifications, including environmental requirements, are commonly used by public clients and frequently criticized by scholars since they distort incentives and decrease service providers’ willingness to be innovative [30]. Adaptability [10,11,26] and incentives [26,30], therefore, seem to be key factors when procuring engineering services to promote carbon reduction.

Williamson [31] reports that (cooperative) adaptability is related to the concepts of high- and low-powered incentives, that the market mechanism is useful when delivering high-powered incentives, and that hierarchies are effective in delivering low-powered incentives. He claims that “low-powered incentives have well-known adaptability advantages” [31] (p. 140), whereas high-powered incentives “impair the ease with which adaptive, sequential adjustments to disturbances are accomplished” [31] (p. 91). In fact, when a client provides a service provider with high-powered incentives, the client will focus on its own revenue (ibid), i.e., cut costs [32], which is not suitable when innovation is needed [31]. Therefore, when designing a contract, creating incentives for appropriate adaptation is important for the client—especially when uncertainties are high [33] and innovation is needed [31].

Adaptability and high- versus low-powered incentives are referred to by several researchers in different procurement stages, e.g., reward system [34,35], bid invitation [36], bid evaluation [37], and specification and evaluation [38]. These researchers either focus on one or two procurement stages, although, as previously stated, the procurement strategy involves control modes within several stages [20,21]. In addition, there is a need for more research on how different control modes affect adaptability [18]. Thus, there is a need to bring together the concepts of high- and low-powered incentives and combine them with the procurement model [20,21], as well as to illustrate their effects on adaptability. By studying public procurement of engineering services to promote carbon reduction from the perspective of organizational control and incentives for adaptability, this study will make two main contributions. First, it will add knowledge to the limited field of procurement of engineering services within construction projects [10,39], where most studies cover the later phases of construction projects and not the design phase [40]. Secondly, it will contribute to the organizational control literature by explaining how different control modes affect adaptability [18].

Although the actions carried out before contract execution are considered important in public procurement [41,42], there are few empirical studies of service procurement covering these phases [43]. Before contract execution, the client has to choose service specification, bid invitation, bid evaluation, and reward system [42,44], so these four procurement stages in the procurement model [20,21] will be in focus.

The purpose of this study is to investigate how public clients may choose control modes and what incentives for adaptability their choice creates for service providers in the procurement of engineering services focusing on carbon reduction.

First, the procurement model developed by Eriksson [20] and Eriksson and Laan [21] will be conceptually adapted from a contractor context to a service context. Secondly, the adapted model will be combined with the concepts of high- and low-powered incentives in a classification model. Thirdly, these two models will be used to analyze the empirical data from a single case study, i.e., an engineering service contract promoting carbon reduction. More specifically, the client’s and the service provider’s perceptions of control modes and their effects on adaptability will be examined.

In this paper there are six sections. The introduction gives some background on the research field and explains the purpose of the study. This is followed by a section describing the theoretical frameworks, including previous research. The third section describes the methods used, and the fourth section presents the empirical data from the interviews and observations. In section five, the analysis is discussed, and concluding remarks are made. In the final section, the theoretical contributions and practical implications of the study are presented.

2. Literature and Frameworks

This section is divided into three parts. In the first part, the control framework (including the three control modes) is presented. In the second part, different activities are connected to the control modes and structured according to the four procurement stages, i.e., the procurement model is outlined. In the third part, the connections between different activities and control modes are tied to the concepts of high- and low-powered incentives, i.e., the classification model is outlined.

2.1. Organizational Control—Three Control Modes

The control framework includes three modes of control, i.e., output, process, and social control, which are based on Ouchi’s [17] market, bureaucracy, and clan control systems, respectively. According to Ouchi [17], the role of organizational control is to “focus on the problems of achieving cooperation among individuals who hold partially divergent objectives” (p. 845). To achieve this cooperation, defining and measuring either the output or the behavior (i.e., the process) is crucial, but if neither is achievable, socialization (clan control) is suggested [17]. Although the framework was developed in an intra-organizational context, it has also been used in inter-organizational contexts [22,45,46].

The three control modes are sometimes classified as either formal (i.e., output and process control) or informal (social control) [22,47]. In output control, only the client specifies and evaluates the output, regardless of the process followed [47,48]. Consequently, output control involves an arm’s length approach, i.e., the “ invisible hand of markets” [49] (p. 758). The client, therefore, assumes that the service provider knows best how to carry out the task [38,50]. In process control, however, the client usually controls the service provider’s process in detail by specifying the process and behavior and then monitoring the execution [51]. Accordingly, it involves a resource-intensive approach [48], characterized by the “visible hand of management” [49] (p. 758).

In contrast to output and process control, social control involves socialization mechanisms to influence the controlled party [47,50]. It relies on the development of common goals, values, and culture to encourage eligible output and behaviors [17,18,22,47]. The intention behind social control is to establish an organizational context where the parties rely on each other and, therefore, where self-control by the service provider is common [22]. Ouchi [17] refers to this common organizational context and culture as a “clan” that orients the service provider towards common goals [19]. Sometimes, researchers also discuss input control, but it is usually not seen as a separate mode, but either belonging to process control [20,52] or social control [17,50]. Depending on the context, input control will here be treated as either process or social control (see Section 2.2.2).

In studies of control modes, some researchers conclude that control has a positive effect on performance [48], whereas others observe a negative effect [23]. For instance, some researchers find that social control is superior to the other control modes in terms of innovation performance [16,28]. Furthermore, the issue of whether or not control modes should be seen as substitutes or complements is widely discussed [18,50]. By arguing for control modes as substitutes, scholars mean that the use of one control mode will reduce the effect of another [17,53], whereas those who argue they are complements mean that using one mode will enhance the effect of another [46,50]. In accordance with Cardinal et al. [18], a holistic view is adopted here, meaning that the client’s control is a mix of different control modes and does not see a single mode as optimal.

2.2. Control Modes in the Procurement of Engineering Services

The model for contractor procurement developed by Eriksson and Eriksson and Laan [20,21] (see

Table 1. Procurement model for clients’ procurement of contractors, adapted from Eriksson and Laan [21].

Procurement Stage	Output Control	Process Control	Social Control
Specification	Spec. by contractor	Spec. by client	Joint spec.
Bid invitation	Open bid procedure	Pre-qual./negotiation	Pre-qual./negotiation
Bid evaluation	Focus on tender price	Focus on authority-based soft parameters	Focus on trust-based soft parameters
Reward system	Fixed-price	Cost reimbursement	Including incentives
Contract formalization	Formal	Formal	Informal
Collaboration	Low usage	Low usage	High usage
Performance evaluation	Output control by client	Process control by client	Self-control by contractor

) serves as a basis for the theoretical framework. The procurement model shows how different alternatives within different procurement stages are tied to different control modes.

Since the procurement model (see Table 1) was originally developed for the procurement of contractors, two changes (shown in italics, see

Table 2. Procurement model for procuring services adapted from Eriksson and Laan [21].

Procurement Stage	Output Control	Process Control	Social Control
Specification	Output-oriented	Process-oriented	Value-oriented
Bid invitation	Open bid procedure	Pre-qual./negotiation	Pre-qual./negotiation
Bid evaluation	Focus on tender price	Focus on authority-based soft parameters	Focus on trust-based soft parameters
Reward system	Fixed-price	Time-and-materials	Including incentives

) have been made to better suit the procurement of services. First, the specification in contractor contracts is related to the project delivery method (i.e., design-build, design-bid-build, and early contractor involvement) [20], which is not applicable to services. Instead, the specification methods (i.e., output-, process- and value-oriented) identified by Axelsson and Wynstra [54] will be used, as they are widely referred to within service procurement [41]. Secondly, due to the fact that time-and-materials compensation is used instead of cost reimbursement in service contracts [55], there is one change to the reward system (process control) as well. Due to the focus here on specification, bid invitation, bid evaluation, and reward system, the other stages are removed.

In the service specification stage, the client specifies the service according to a specification method [54], which allocates responsibility for the quality of the results [20]. In the bid invitation stage, the client chooses which procurement procedure to use [20], so the client also decides the number of bidders allowed and whether there will be any negotiations. In the bid evaluation stage, the client formulates and evaluates the awarding criteria, i.e., the extent to which price and/or quality will be the decisive factors in determining the winning bid (ibid). In the reward system stage, the matter of risk and responsibilities between the parties is central [56]. Specification and reward system are therefore related to responsibility (quality and financial), whereas bid invitation and bid evaluation are related to the selection. The content of Table 2, i.e., one of the two main models used in the analysis, is explained below.

2.2.1. Specification

In output control, output-oriented service specification is used, which means that the service provider translates the functions specified by the client into activities [54], which in turn makes the service provider responsible for quality [20]. In process control, process-oriented service specification is used, and the client specifies in detail what the service provider should do [54], which means the client is responsible for quality [20]. In value-oriented specifications (also called outcome-oriented), the client describes its needs or what value the service should fulfill [54]. After that, the parties jointly develop the activities needed to achieve that and thus share the responsibility for quality [20]. In both output- and value-oriented specifications, it may be difficult for the parties to interpret the output/value similarly, which is typically the case when the service is not simple, and the output/value is difficult to measure [54].

2.2.2. Bid Invitation and Bid Evaluation

In an open procedure, the client invites a large number of bidders to compete [20], thus adopting a rather short-term perspective [52]. Similarly, using the lowest price in the bid evaluation also implies a short-term perspective [52], as the client does not utilize the opportunity to influence the service provider [20] but adopts a more laissez-faire approach [52]. Both open procedure and lowest price are therefore considered to be tied to output control [20]. In contrast, by using pre-qualification and negotiation, a limited number of bidders are involved [20], and the bids are up for negotiation [57], which is a more long-term perspective [52]. Process control is closely related to a long-term perspective, as there are fewer incentives to prioritize immediate pay-offs [52]. In addition, social control involves socially investing in the partner, which is enhanced by long-term relationships [38]. Negotiations with a limited number of bidders facilitate long-term relationships and thus are related both to process and social control [20]. Due to the importance of the long-term perspective, using the best price–quality ratio in the bid evaluation is connected to process and social control [21]. In process control, the focus is more on authority-based soft parameters (ibid), including input control measures such as experience and competence [46], whereas, in social control, the quality aspect is trust-based [21], which includes input control measures such as innovation ability [16], collaborative ability, and shared values [20].

2.2.3. Reward System

In output control, the service provider is rewarded for its output, typically with a fixed-price contract providing a fixed payment from the client [20]. Fixed-price contracts are usually used for simple services, where the output is easily specified [35]. In contrast, when services are more complex and uncertain, the service provider is usually compensated using a time-and-materials contract, which is based on time and resources (i.e., price per consultant, per hour) [20]. In comparison to fixed-price contracts, the specifications do not have to be as precise since the client bears the financial risks, and changes in the contract execution phase are easier to carry out [35]. In social control, contracts, including incentives, can be used to improve quality and reduce costs [20], e.g., combining a time-and-materials contract with bonuses linked to non-economic factors such as quality, timeliness, and environmental impact [58]. Another example could be an incentive-based contract including two phases, where the first phase is compensated on a time-and-materials basis and the second on a target cost basis plus a pains/gains sharing between the parties if the real cost differs [58].

2.3. High- and Low-Powered Incentives in the Procurement of Engineering Services

In accordance with Williamson [31], high- and low-powered incentives will be used to conceptualize adaptability. Williamson [59] asserts that the market mode, which is tied to output control [20] through the “invisible hand of markets” [45] (p. 758), is characterized by high-powered incentives and poorly suits cooperative adaptation. In contrast, the hierarchy, which is tied to process control [20] through the “visible hand of management” [45] (p. 758), is characterized by low-powered incentives and supports cooperative adaptation [59]. The hybrid, which is tied to social control [20], also supports cooperative adaptation [59]. High-powered incentives, therefore, lead to low cooperative adaptability and low-powered incentives lead to high cooperative adaptability [32,59].

The reason why cooperative adaptation is facilitated by hierarchy is that, within an organization, it can be exercised by giving orders without renegotiations. In contrast, in a market setting, two organizations are involved that both have different profit streams, so adaptations are complicated, and renegotiations are needed. In order for a market setting to facilitate adaptation, it has to be incorporated into the contract from the beginning through, for example, time-and-materials compensation [31]. Even though the logic above is based on either intra- or inter-organizational contexts, both high- and low-powered incentives are referred to in purely inter-organizational settings [35,37]. Below, the concepts of high-and low-powered incentives are tied to the procurement model (see Table 2) and structured accordingly.

2.3.1. Specification

Aulakh and Gencturk [38] conclude that when the client uses process or social control, the service provider is more willing to comply with changes to the specifications and new directives from the client, and there is thus a positive impact on adaptability. In contrast, they report a negative impact on adaptability when using output control due to the service provider being responsible for the quality, as well as bearing the financial risk in output control, but not in process or social control.

Thus, process- and value-oriented specification are related to low-powered incentives (i.e., high adaptability), especially if the client has enough understanding and information, and output-oriented specification is related to high-powered incentives (i.e., low adaptability). However, according to Tadelis [32], the service provider’s incentives (both high- and low-powered) are more closely connected to the reward system than to the specification method.

2.3.2. Bid Invitation and Bid Evaluation

Bajari et al. [36] state that when contracts are complex, specifications are incomplete, and the client lacks understanding of the tasks, negotiations are favored because negotiations facilitate adaptation and are able to handle uncertainties and complexities. They also conclude that open procedure does not provide any incentives for adaptability but only for cutting costs [36,52], i.e., high-powered incentives. In contrast, by using pre-qualification and negotiated procedures, the service providers’ incentives will be long-term [52] and facilitate adaptability [36], i.e., low-powered incentives.

In regard to bid evaluation, using the lowest price creates no incentives for adaptation [60] or higher quality [37] and has a short-term focus on cost benefits, i.e., high-powered incentives [37,58]. Using the best price-quality ratio, however, provides incentives for adaptation [60] and for higher quality as long as the quality aspects are well defined and verifiable in the contract execution phase, i.e., low-powered incentives [37].

2.3.3. Reward System

In regard to reward system, fixed-price contracts provide incentives to reduce costs but result in high friction when changes are needed (i.e., high-powered incentives) [35]. In contrast, time-and-materials contracts better accommodate changes but lack incentivizing mechanisms to reduce costs (i.e., low-powered incentives) (ibid). Fixed-price contracts, therefore, are not suitable when adaptability is needed, whereas time-and-materials compensation is [36,61]. With time-and-materials contracts, the service provider has incentives to provide high-level competencies [55,56] and high-quality services compared to fixed-price contracts, which incentivize only cost-cutting [34,36].

The classification of fixed-price and time-and-materials compensation is rather straightforward, but the classification of contracts, including financial incentives, can be more ambiguous. In contracts including time-and-materials compensation, plus an incentive mechanism such as a bonus, the service provider has an incentive to reach the bonus [58] and improve quality [20], meaning that the incentive could be considered low-powered. On the other hand, in an incentive-based contract, adjusting the target cost tends to be a rather complex matter that causes friction between the parties (similar to fixed-price contracts [35]), and thus it could be considered to provide high-powered incentives.

2.3.4. Classification Model

In

Table 3. Classification of incentives in public procurement of services.

Procurement Stage	Output Control	Process Control	Social Control
Specification	Output-oriented (hp)	Process-oriented (lp)	Value-oriented (lp)
Bid invitation	Open bid procedure (hp)	Pre-qual./negotiation (lp)	Pre-qual./negotiation (lp)
Bid evaluation	Focus on tender price (hp)	Focus on authority-based soft parameters (lp)	Focus on trust-based soft parameters (lp)
Reward system	Fixed-price (hp)	Time-and-materials (lp)	Including incentives (hp/lp)

below, the classification of incentives is added to the procurement model for services (see Table 2). Alternatives that are considered to provide low-powered incentives are marked with (lp), and high-powered incentives are marked with (hp). This model is used in the analysis to classify the client’s choice of procurement alternatives as providing either high- or low-powered incentives for the service provider.

3. Method

A qualitative and explanatory approach was adopted for this study based on one case, i.e., an engineering service contract focusing on carbon reduction within the Swedish Transport Administration (STA). A single case study serves as “a very powerful example” [62] (p. 20) that richly describes a certain phenomenon [62,63]. A case study can be used to generalize to other situations, i.e., “analytic generalization” [64] (p. 18), so the findings can build theoretical premises that work as assertions in situations other than the one being studied (ibid). The STA is the governmental agency responsible for long-term infrastructure planning and for the procurement of construction and maintenance works on state-owned roads and railways. It is the largest client of infrastructure investments in Sweden, so its engineering service contracts have a huge impact on other public clients.

3.1. Case Selection and Description

The engineering contract studied was part of a pilot project designed to reduce carbon emissions by 50% in the construction phase, compared to a standard project. It was selected based on its suitability [63] to illustrate the phenomenon, i.e., the public procurement of engineering services that promote carbon reduction. It was a road infrastructure project, and the cost of the engineering services was approx. EUR 2–2.5 million. There were two phases in the contract: physical planning and preparation of tender documents for the contractor.

The physical planning phase included various investigations and consultations with stakeholders and an environmental impact assessment. This phase resulted in a final plan proposal (in this case, a road plan) that needed to be approved by the Country Administrative Board. When the plan was approved, the STA got legal permission to acquire the land needed for the construction work. In the tender document phase, the contracted engineering consultancy company (ECC) provided the STA with the tender documents to be used in the procurement of the contractor (design-bid-build). Thus, the ECC provided the design for the contractor to use in the construction phase.

3.2. Data Collection

As with most case studies, the intention was to combine different sources of empirical evidence (triangulation) in order to verify the data [65], so the interviews (primary data) were complemented by observations (primary data) and document studies (secondary data).

According to Alvesson [66], interviews are one of the best and most common ways to collect data in qualitative research. In order to gain a comprehensive view of the client’s and service provider’s perception of the contract, 16 respondents (10 from the client and 6 from the service provider) were selected for interview (see

Table 4. List of respondents.

Working Role	Organization	Length (min)
Project manager	Client	125
Project engineer	Client	105
Purchaser	Client	125
Innovation strategist	Client	60
Unit manager	Client	45
Specialist in environment	Client	110
Specialist in bridge construction	Client	110
Specialist in road design	Client	110
Specialist in road equipment	Client	45
Specialist in geotechnics	Client	75
Bid manager	Service provider	90
Assignment manager	Service provider	90
Design manager	Service provider	90
Climate coordinator	Service provider	80
Plan coordinator	Service provider	80
Function coordinator	Service provider	75

).

The respondents were selected according to their suitability, i.e., using theoretical sampling [63]. At the STA, the interviews were held with the people most involved in the procurement preparation. At the winning ECC, the respondents were those people most involved in the bidding process as well as at the beginning of the contract, i.e., the project management team. All the interviews were semi-structured to enable engagement from both the researcher and the respondent and lasted for 45–125 min. The questions were inspired by the procurement model [20,21] and the control modes [17]. Due to COVID-19, all the interviews were carried out online. All interviews were recorded and transcribed with the consent of the respondents.

The observations were carried out by the main author during the procurement preparation and at the contract kick-off meeting in order to capture the (real-time) dialogue. The researcher acted as a “complete observer” and thus did not participate at all in the dialogue [67]. Due to COVID-19, all the meetings were held online, so the researcher could not capture any other aspects than the dialogue. The researcher was able to capture most of the discussions by taking notes. All the observations are outlined in Appendix A, including the participants and the length of the meetings.

3.3. Data Analysis

Siggelkow [62] asserts that case studies are useful in three ways: inspiration, illustration, and motivation. By this, he means that there are shortcomings inherent in purely conceptual contributions and that case studies can be used to illustrate phenomena in real-life settings. In line with this, the empirical data should first be used to illustrate the procurement model, showing how a client may choose control modes when procuring engineering services to promote carbon reduction, and thus why the model could be an important tool to analyze that. Secondly, the empirical data should be used to illustrate the usefulness of the classification model, showing how to create incentives for adaptability for a service provider to promote carbon reduction, and thus why the model could be an important tool to analyze that.

A flexible pattern matching approach was used to analyze the transcribed interviews and the notes from the observations, which is a recommended approach to expand and test existing theories [68]. Throughout the data analysis, therefore, the observed empirical patterns were matched with the theoretical patterns to make interpretations of the data. In addition, to make sense of the data, tables were created (i.e., visual mapping) [69]. In order to strengthen validity [70] and robustness [62], throughout the analysis, the researchers kept asking “why” when similarities and differences appeared between the theoretical patterns and empirical data.

The analysis was mainly carried out in two steps. First, the empirical data were compared to the procurement model for services (see Table 2) to identify the control modes in use. Secondly, the empirically identified control modes were matched with the classification model (see Table 3) to identify their prescribed effects on adaptability.

4. Findings

This chapter begins by explaining the aim of reducing carbon emissions (i.e., the context), and then the findings are presented in the same way as the procurement model and the classification model.

4.1. Reducing Carbon Emissions

Several of the respondents from the STA indicated that, since the engineering contract was part of a pilot project, it was right to drive the development towards carbon emission reductions using that contract and other contracts. They believed that the ECC should promote carbon reduction through solutions and conditions for the contractor and measure compliance. Respondent STAa hoped the contract would “result in something that the entire industry could utilize in the future.” Since it was a contract within a pilot project aiming at carbon reduction, several respondents from both parties said that many people were interested in the work within both organizations and also within other ECCs. Respondent ECCk said that “to help society reduce the climate footprint is totally in line with what we want [within our company]. Internally, it has generated enormous interest.” Some of the respondents from the ECC also expressed that, since this was a pilot, it might become the standard way of working in the near future, and thus it was an important contract for them.

4.2. The Choice of Procurement-Related Control Modes and Their Effects on Adaptability

In each section below, there is a description of the empirical findings and a table visualizing some of the citations linked to the three control modes.

4.2.1. Service Specification

The respondents said that, in the service specification, the STA described the output of the physical planning and tender documents by referring to several documents the ECC should develop. As visualized in

Table 5. Citations regarding the specification stage.

Output Control	Process Control	Social Control
“It is clearly described which memorandums, programs and documents we want” (STAa).	“In general, our regulation is very detailed. We have a lot of requirements” (STAd).	“It is preferred if they first investigate and then we decide how we continue” (STAd).
“It is pretty clear what we should do. How we do it is up to us” (ECCm).		“It is good that we can adapt depending on what is actually demanded” (ECCo).

, respondents from both parties seemed to know which documents the ECC would be handing in and what these should include. However, it was up to the ECC how it would work, and it seemed like the client was partly using output control.

Most of the respondents at the STA argued that the specification, in terms of possible technical solutions, was rather flexible and that there would always be adaptations to the specifications in the contract execution phase, as there were many uncertainties, which was agreed by respondents from the ECC. On the other hand, the technical solutions also had to obey the requirement regulations, which they considered rather rigid. Therefore, several respondents said that there would probably have to be dispensations from the regulations. For example, Respondent ECCk said that:

“There will be occasions when we have to tell the STA that we can do it this way and decrease the carbon emissions by this amount, but then we will have to get dispensations from the regulations. (…) The STA have to be aware of this.”

Since the specialists at the STA were responsible for several requirement regulations, they needed to assist in these adaptations. Hence, it seemed like the client was using output control in combination with process control. Due to the rigid regulations that had to be obeyed, process control seemed to be the dominant control mode.

4.2.2. Bid Invitation and Bid Evaluation

The bid invitation was carried out as an open procedure. In some of the observed meetings, the participants from the STA said they were certain about what they wanted the ECCs to submit in their bid, i.e., a description of contract execution. Thus, they saw no reason to use negotiations, which were more adaptable. Similarly, one of the respondents from the ECC said that the choice of procurement procedure did not really matter to him, while the others said they lacked competence in procurement procedures.

Many of the respondents from the STA said that, in general, the internal specialists within the project organization were very experienced and had a high competence. However, when it came to this pilot project requiring climate-friendly solutions, most of them believed they lacked understanding. Therefore, some of the respondents from the STA thought it very important that the winning ECC should be the best, most engaged, and most competent in terms of promoting carbon reduction. Based on that, the best price-quality ratio was argued to be the natural choice, meaning that the bidders should describe the competencies and way of working they intended to use to reach the carbon emissions target and how this should be verified (i.e., social control/self-control).

Because of the evaluation method, the respondents at the ECC said they realized the carbon reduction target was important, and thus they were able to adapt their bid accordingly. For instance, they put a climate expert in their project management team and added an innovation mentor to their support function, which was unusual for them. Furthermore, the respondents also suggested a new iterative and innovative way of working in the description of the contract execution (see

Table 6. Citations regarding the bid invitation and bid evaluation stages.

Output Control	Process Control	Social Control
“We could not find what exactly to negotiate, and thus we chose not to.” (STAa).	“Are we able to put requirements on the ECCs to have competence within production?” (STAa).	“We would like them to present their organization and what way of working they intend to use in order to reach the carbon reduction goal.” (STAb).
		“We believed that the [climate] issue was important and therefore we formed this organization in the way we did.” (ECCl).

).

4.2.3. Reward System

Due to the many uncertainties and the need to facilitate innovation, it seemed clear to most respondents from the STA and the ECC that time-and-materials was the most appropriate reward system for this contract (see

Table 7. Citations regarding the reward system stage.

Output Control	Process Control	Social Control
“Fixed-price would not provide the right pre-requisites for the contract. (…) Having this demanding goal, we cannot put that much [financial] risk on the engineering consultants and I do not think we would reach the goal then.” (STAi).	“I think time-and-materials is the right choice when you want some new thinking.” (STAj).	“It crossed my mind that there seldom is an incentive in the engineering service contract, since the effects do not show until the production phase.” (STAb).
“I think we would have been very uncertain and very unwilling to submit a bid if it had been a fixed-price, due to there being a lot of uncertainties along the road.” (ECCk).	“I think it [time-and-materials] was a pre-requisite for us to submit a good bid and for us to jointly reach the [reduction] goal.” (ECCk).	“It would be nice if you could get a part of the carbon reduction as a bonus.” (ECC1).

). The respondents from the STA said there were great uncertainties that would probably lead to a lot of discussions and reworking throughout the contract. By this, they meant that a time-and-materials contract would add adaptability and cause less friction since the ECC’s costs would still be covered. Similarly, the respondents from the ECC said that time-and-materials compensation would give them greater possibilities to be adaptable and carry out more relevant investigations due to their costs being covered. However, some of the respondents from both parties were uncertain whether the STA would allow the ECC to work as many hours as it believed it needed. Furthermore, the respondents from the ECC said that the choice of reward system would also make it possible for them to use their most suitable and high-level competencies.

5. Discussion

This discussion is divided into two sections. The first focuses on classification according to previous research (see Table 3), and the second on the perceptions of the client and the service provider.

5.1. Suggested Levels of Incentives

This section will begin with a summary of the empirically chosen control modes and their addition to the procurement model for services, see

Table 8. The client’s choice of control modes.

Procurement Stage	Alternative	Control Mode
Specification	Output/process-oriented	Output/process control
Bid invitation	Open bid procedure	Output control
Bid evaluation	Focus on trust- and sustainability-based soft parameters	Social control
Reward system	Time-and-materials	Process control

. Secondly, the empirically chosen control modes will be combined with the classification model (see Table 3). This discussion will be based on previous research.

The client’s choice implies that it took responsibility for quality [20] and bore the financial risk [35]. In addition, due to the client’s perceived lack of understanding of carbon reduction, the client seemed keen on selecting a service provider that had a superior understanding. Therefore, it aimed for a service provider that shared the same values [20]. In addition, the client considered it important for the service provider to have the competencies and working processes to be able to lead innovation [16], which the client believed would be necessary in the contract execution phase.

According to previous research, the client’s choice of output- and process-oriented specifications influences the service provider’s willingness to adapt to changes both negatively and positively [38]. However, due to the strong requirement regulations, the specification here is classified as low-powered. In regard to bid invitation, the client’s choice influences cost-minimizing rather than incentives to improve quality [36,52], so its choice is classified as high-powered. In line with previous research, the client’s choice of best price-quality ratio in the bid evaluation stage should have a positive effect on the service provider’s incentives to deliver high quality [37], i.e., their choice is classified as low-powered. In regard to reward system, the client’s choice of time-and-materials compensation has a positive effect on adaptability [35] and on the service provider’s incentives to deliver high quality [34,35,55,61], so the client’s choice is classified as low-powered. The discussion above is summarized in

Table 9. The client’s choice of control modes and their classification regarding incentives.

Procurement Stage	Alternative	Control Mode	Incentives
Specification	Output/process-oriented	Output/process control	Low-powered
Bid invitation	Open bid procedure	Output control	High-powered
Bid evaluation	Focus on trust- and sustainability-based soft parameters	Social control	Low-powered
Reward system	Time-and -materials	Process control	Low-powered

.

The findings presented in Table 9 indicate that the client’s choice seemed to provide more low-powered than high-powered incentives. According to Tadelis [32], the incentive provided by the reward system is more important than the one provided by the specification method, and it thus determines whether the client will provide high- or low-powered incentives. However, in this case, both are classified as low-powered. When it comes to bid invitation and bid evaluation, both high- and low-powered incentives seemed to be provided. Due to the client’s choice of trust- and sustainability-based soft parameters, it is reasonable to argue that the service provider had a low-powered incentive to provide high-level competencies and suitable working methods, no matter what bid invitation was chosen. This is simply because a service provider offering higher quality is more likely to win the contract. However, a negotiated procedure could clarify uncertainties in the bid or improve the bid [57], which could potentially make the low-powered incentives to reduce carbon emissions even stronger. Indeed, the procurement strategy as a whole seems to influence the service provider’s low-powered incentives to promote carbon reduction.

5.2. Perceived Levels of Incentives

The empirical data suggest that both the client and service provider felt that the choice of process (and output) control in the specification positively influenced the service provider’s adaptability and thus served as low-powered incentives [31]. However, due to the demanding carbon reduction targets and the fact that the service provider was thought to have a greater understanding than the client, it is uncertain whether the adaptability should have been initiated by the client (i.e., process control) rather than jointly (i.e., social control).

In regard to bid invitation, the empirical data suggest that neither the client nor the service provider perceived the choice of procurement procedure as important in terms of the incentives created for the service provider, which is different from previous studies [36,57]. In contrast, the client and the service provider appeared to believe that the bid evaluation had a greater influence on the incentives than on the procurement procedure. The empirical data show that the client perceived it was important to select an engaged service provider, i.e., a service provider that shared the same values and goals, similar to social control [17,18,22]. The focus on trust- and sustainability-based soft parameters seemed to signal to the service provider that reducing carbon emissions was an essential part of the contract. The client’s choice of control mode, therefore, seemed to influence the service provider to select high-level competencies (i.e., a climate expert and innovation mentor), as well as a more innovation-friendly working process, in order to be competitive. Thus, in accordance with previous research, the client seemed to provide low-powered incentives [37].

In regard to reward system, the empirical data show that the client believed time-and-materials compensation provided low-powered incentives for the service provider to reduce carbon, as suggested by Bajari and Tadelis [35]. Similarly, respondents from the service provider said that it would allow them to be adaptable and choose their most suitable (i.e., high-level) competencies, which is also suggested by previous research [55,56,61].

In summary, allowing the service provider greater adaptability (i.e., low-powered incentives) seemed more important than high-powered incentives for both parties when procuring engineering services aimed at carbon reduction. The empirical data indicate that the client’s choice of control modes provided low-powered incentives, both through the selection (bid invitation and bid evaluation) and in relation to responsibility (specification and reward system), thus confirming and extending the argument by Bajari et al. [36] to include more procurement stages. Although previous research suggests that open procedures provide high-powered incentives for the service provider [57], the empirical data show that the bid evaluation offset the effect of the high-powered incentives, turning them into low-powered incentives instead. Thus, the procurement strategy as a whole was perceived to provide low-powered incentives for the service provider to promote carbon reduction in the construction phase, both in the selection and in regard to responsibility.

6. Conclusions

In this study, a procurement model for services (see Table 2), based on an existing procurement model by Eriksson [20] and Eriksson and Laan [21], was developed. Based on the developed model and the concepts of high- and low-powered incentives, a conceptual classification model (see Table 3) was created to illustrate how different procurement choices influence high- and low-powered incentives for the service provider. The classification model was used to analyze the client’s and service provider’s perception of adaptability, showing that low-powered incentives dominated the procurement-related control modes overall.

6.1. Theoretical Contributions

The client thought that a combination of different control modes was suitable when procuring engineering services aiming for carbon emission reduction in the construction phase, i.e., process (service specification, reward system), output (service specification, bid invitation), and social control (bid evaluation). The client believed that it should be responsible for the quality and bear the financial risk, while it was important to award the service provider with high-quality competence and innovative working processes because the client seemed to lack understanding of how to reduce carbon emissions. This knowledge contributes to the research on procurement strategies aiming for the reduction of carbon emissions [13,14], as well as to the control literature in general.

Scholars have hitherto studied a maximum of two of the procurement stages (i.e., specification, bid invitation, bid evaluation, and reward system) and their influence on incentives [36]. However, they have neglected the fact that a procurement strategy consists of procurement-related control modes in more than two stages. This study adds to previous research, showing that it is important for a procurement strategy to include activities within several procurement stages [20,21]. The main theoretical contribution of this study is the conceptual classification model that was developed, which makes it possible to illustrate how a client’s choice of procurement-related control modes may affect adaptability. Adding to the literature on high- and low-powered incentives [32,36,37], it can be concluded that it is important to focus on incentives both in the selection and in relation to responsibility—especially in the bid evaluation and reward system stages. Thus, the classification model may serve as an important tool for visualizing incentives within different procurement stages.

Another contribution of this study is that, when procuring engineering services to promote carbon reduction in the construction phase, low-powered incentives are perceived as superior to high-powered incentives. This is because the service provider is then incentivized to select consultants with higher competence and work in a more iterative and innovative working process. This is an important contribution to the literature on control modes’ effect on adaptability [36,37], as well as to the research on procurement strategies aimed at reducing carbon emissions [13,14].

6.2. Practical Implications

Due to the key role of engineering services in solving emerging issues related to carbon reduction [9,27], the findings of this study are important for public clients. In order for them to meet their carbon reduction targets, procurement strategies are of great importance. However, it is difficult for public clients to choose these appropriately [13]. This study contributes important knowledge for practitioners, as it demonstrates the usefulness of seeing the procurement strategy as a whole, including control modes within the stages of service specification, bid invitation, bid evaluation, and reward system. In addition, the control modes’ influence on incentives should also be considered as a whole. The classification model could therefore serve as a practical tool to increase much-needed knowledge on incentives [29].

6.3. Limitations and Further Research

Due to its focus on the phases before contract execution, this study is unable to capture potential consequences that will appear later on in the contract execution phase. Therefore, longitudinal studies including the contract execution phase may be of considerable benefit. In this study, the main focus was on control modes and adaptability in the procurement of engineering services promoting carbon reduction in the construction phase. However, this study also indicates that innovation is important in reaching climate-related targets, and thus there is a need for further studies to explore broader perspectives on innovation and sustainable development.