Contractual learning and organizational design in public procurement Open Access

Public organizations face increasing pressure to use resources efficiently when procuring goods and services (Schapper et al., 2006; Patrucco et al., 2019; Thai, 2001). This reinforces the need for effective contractual governance in public procurement relationships. According to transaction cost economics (TCE), organizations strive to mitigate supplier opportunism by choosing contractual safeguards that match the level of asset specificity, uncertainty and transaction frequency (Williamson, 1985). However, TCE does not explain how contractual arrangements evolve over time through cumulative experience and external feedback (Nooteboom, 1992). Hence, drawing on evolutionary economics (Nelson and Winter, 1982) and capability-based theories (Zollo and Winter, 2002), scholars have proposed that organizations progressively enhance their governance of contractual relationships by developing contracting capabilities. These capabilities are grounded in the specialized knowledge of technical, commercial and legal personnel within the organization (Argyres and Mayer, 2007). The gradual refinement of these capabilities through accumulated organizational experience is known as contractual learning. Prior research in the private sector characterizes contractual learning as experiential and contract-specific, shaped by prior interorganizational ties and internal firm characteristics (Arino et al., 2014; Faems et al., 2008; Hallberg, 2024; Lumineau et al., 2011; Mayer and Argyres, 2004; Vanneste and Puranam, 2010; Xing et al., 2021).

Research on contractual learning in public organizations remains relatively scarce despite the fact the legal and regulatory requirements in the public sector are likely to give rise to a different set of contractual problems and solutions than in the private sector (Knutsson and Thomasson, 2013). For example, by making it more difficult to use informal, relational mechanisms and agile contract revisions (Lumineau et al., 2011). This raises important questions concerning how organizations in the public sector adapt to the additional regulatory constraints placed on them. Given the special demands placed on public organizations, it is likely that they will develop novel organizational approaches to handle regulatory pressures and that they may follow different learning trajectories than private firms (Argote, Lee and Park, 2021). Hence, based on previous research indicating that internal organization play an important role in facilitating contractual learning (Hallberg, 2024), we specifically seek to answer the following research question: How does organization influence contractual learning in public procurement? To explore this question, we draw on two case studies of a municipality’s procurement of cleaning services and a research organization’s procurement of a technical system. Our findings suggest that integrating legal specialists early in the procurement process supports the design of contractual templates and safeguards, while technical experts play a vital role in providing experiential feedback on contract performance during the governance phase. We make three contributions to the studies on contractual learning and public procurement. First, our findings indicate that contracting is particularly challenging in public procurement because of the reduced adaptiveness resulting from public procurement regulations and that public organizations respond to these conditions by improving contractual safeguards and outcome-based evaluation criteria (see Girth and Lopez, 2019; Kleemann et al., 2012). Second, we also contribute to research on the role of specialization in learning to design and evaluate procurement contracts. For example, previous research describes the distinctive roles of managers, engineers and lawyer in the contractual process (Argyres and Mayer, 2007). We contribute to this literature by demonstrating how functional specialization in public procurement is shaped by decisions to centralize and consolidate procurement activities. Finally, we contribute to research on the impact of structural integration on learning (Jansen et al., 2009) by outlining new mechanisms concerning how specific forms of integration facilitate contractual learning in public procurement.

A key proposition of TCE is that organizations mitigate contractual hazards by matching governance structures with transactions depending on how the transactions score on attributes such as asset specificity, uncertainty and frequency (Williamson, 1985). Transactions subject to greater hazards and risk of opportunism because of high levels of asset specificity and uncertainty will, thus, be matched with governance structures that offer greater levels of control (e.g. extensive safeguards). However, it is likely that organizations will learn from prior experience and gradually get better at aligning the contractual design with the attributes of the transaction (Tsai and Chiu, 2024). These dynamic processes stem from the accumulation of new experiences, trial-and-error processes and the routinization of successful behaviors, which when properly articulated and codified in the organization, may become the foundation for organizational capabilities that allow the organization to produce desired outcomes in a stable manner (Argote and Miron-Spektor, 2011; Zollo and Winter, 2002).

Previous research show that contractual learning processes work incrementally to locally improve contracts in a specific relationship (Mayer and Argyres, 2004). Learning is typically stronger for technical than for legal terms (Vanneste and Puranam, 2010), dependent on relational dynamics and bargaining power (Faems et al., 2008), and involves learning about the contractual process, the parties and the transaction (Lumineau et al., 2011). Overall, contractual learning processes have been described as highly experiential (Arino et al., 2014; Xing et al., 2021). Contractual learning involves the development of knowledge at the organizational level that is dispersed across different employees, such as economic knowledge held by managers and commercially oriented employees, technical knowledge held by engineers and technical personnel and legal knowledge held by lawyers and internal legal counsels (Argyres and Mayer, 2007). An effective procurement organization, which allows the organization to learn from previous experiences, will thus likely involve the specialization of employees within these three functions, as well as mechanisms for coordinating tasks performed in different specialized units.

Structural differentiation and integration: The structural differentiation of different units within an organization allows each unit to adapt more effectively to the demands placed on it by the environment given the tasks performed within the unit (Lawrence and Lorsch, 1967; Tushman and Nadler, 1978). For example, in the context of public procurement, the economic function may refer to units that are specialized on issues related to the economic performance of contracts entered by the organization (e.g. procurement officer); the technical function refers to units that are specialized on matters related to the functionality of the product or service being procured (e.g. a technical department); and the legal function refers to units that are specialized on the legality, verifiability and enforceability of contractual commitments (e.g. administrative lawyer). These three functions play different roles in the contractual process: The employees in the economic and technical functions are likely to hold knowledge concerning the development of contractual terms related to the parties’ roles/responsibilities and their communication, while the legal function is an important knowledge repository concerning decision and control rights, dispute resolution and contingency planning (Argyres and Mayer, 2007). We build on these different types of terms and clauses in procurement contracts by distinguishing between coordination- and control-oriented terms (Mellewigt et al., 2007; Nooteboom et al., 1997). This distinction allows us to explore how distinctive forms of organization may influence different types of learning in public procurement.

Addressing complex problems in organizations, such as the design and governance of a public procurement contract, often requires the alignment and coordination of knowledge residing in different departments and units. Organizational theorists have defined this form of structural integration as “the process of achieving unity of effort among the various subsystems in the accomplishment of the organization’s task” (Lawrence and Lorsch, 1967:4). Organizations are, thus, often required to set up differentiated units that enable task specialization, while also considering how the involved units may be structurally integrated to enable coordination between specialized tasks (March and Simon, 1958). Such integration mechanisms may, for example, involve the design of cross-functional interfaces, the set-up of performance evaluation systems and facilitating social cohesion between members of different units (Jansen et al., 2009). In the context of public procurement, this highlights several important empirical questions concerning how specialized employees are organized and how organizational integration mechanisms across different units affect contractual learning.

Public procurement typically imposes constraints on procuring organizations, such as transparency, accountability, impartiality and political oversight, along with legal limitations (Schapper et al., 2006). These rules aim to secure democratic influence over public resources, foster fair competition among suppliers and prevent corruption. At the same time, they can give rise to additional costs (Johnson et al., 2003), for instance when multiple goals (e.g. economic, technical and societal) must be balanced and the flexibility to modify or terminate contracts is restricted. These differences from private procurement are likely to affect both how the procurement function is organized and how contractual learning processes unfold. Previous studies show that public procurement can be structured in widely varying ways. Patrucco et al. (2019), for example, identify six organizational configurations in municipalities, ranging from a fully centralized procurement department to decentralized systems where individual departments hold decision-making authority. Similarly, McCue and Pitzer (2000) find that many US Government procurement organizations use a mix of centralized and decentralized structures. Centralization and the allocation of authority influence how goals are perceived (Glas et al., 2017), how stakeholder interests are incorporated (Kamann, 2007), how professional competencies evolve (McVitt et al., 2012) and whether organizations can leverage cross-functional expertise (Schiele, 2005). Alongside these structural variations, prior research has also examined a range of public-sector contract designs from behavior- to performance- and outcome-based contracts (Kleemann et al., 2012; Girth and Lopez, 2019; Ng et al., 2013) that address different policy and organizational objectives. In this paper, however, we primarily focus on how internal organization influence contractual learning processes, rather than on comparing distinct contract types in public procurement.

This study uses an inductive case study design (Eisenhardt, 1989; Yin, 2018) of two organizations engaged in public procurement. Respondents and the organizations are presented using the pseudonyms. MunOrg is a municipality, and the case study is focused on MunOrg’s procurement of cleaning services in the period of 2012–2020 and the organizational learning associated with this activity. ResOrg is a research organization, and the case study is focused on a larger procurement project of an advanced technical system between 2012 and 2020. The two cases in the study were selected following an inquiry into suitable public organizations where advice from third parties set researchers in contact with relevant decision-makers in the two organizations. Explorative interviews were set up to evaluate the suitability of the cases. The result of the explorative interviews indicated that the two selected cases complemented each other. MunOrg’s procurement of cleaning services presented an opportunity to study procurement of high-volume services that had historically presented significant challenges related to contractual design and governance. ResOrg’ procurement of a technical system, on the other hand, presented an opportunity to study a large procurement project for technically advanced equipment, which had also been subject to contractual renegotiations and organizational changes. Hence, the different conditions in the cases were expected to give rise to different contractual solutions and challenges, which in turn can be expected to affect contractual learning (Yin, 2018).

Data collection for the case-studies was done through semi-structured interviews and the analysis of internal documents. Data collection followed a case-study protocol that outlined data collection procedures, data sources and the specific questions the collected data should answer (Yin, 2018). The protocol included general questions concerning the organization and its external environment and more detailed questions related to the organizational units and processes associated with procurement, the structure of procurement contracts and instances of learning. The 21 semi-structured interviews were conducted. These interviews were mainly held at each site and had an average duration of approximately 1 h. In addition, internal documents also served as a data source. These documents included procurement contracts with suppliers, contractual revisions and documentation of the procurement process. Details about the cases and the collected data are presented in Table 1.

The data analysis was guided by case study research methods (Yin, 2018; Eisenhardt, 1989) and used the method of constant comparison (Glaser and Strauss, 1967; Strauss and Corbin, 1998). The interviews were transcribed and closely read. Emerging themes were noted in a separate document while marking the text in the transcript that the theme referred to. Through iterative comparison of the text belonging to different themes, the set of themes was refined so that a comprehensive categorization was accomplished. Consistent with the principles of open coding, themes were then described to capture the content of the quotations included under each theme. A case description was written based on this thematic structure. Each case description was structured according to the following themes: the transacting parties, technological conditions of the transaction, contract design, contracting organization and processes and learning outcomes (Tables 1 and 2). Once the cases had been compiled according to this structure, a comparison was made between the cases to identify theoretical patterns related to procurement contracting (cross-case analysis). During this analysis, it became clear that internal organization played an important role in promoting contractual learning. Subsequent empirical analysis was, thus, focused on this aspect of organizational learning. The last phase of data analysis concerned a comparison between the empirical pattern emerging from the cross-case analysis and the pattern found in previous research on public procurement and organizational learning (Yin, 2018). We used concepts from previous theory and research to refine empirical categories. Organizational learning is, thus, defined as “a change in the organization that occurs as the organization acquires experience” (Argote and Miron-Spektor, 2011: 1124), which, in our study, involved mapping changes in procurement contracts, as well as other knowledge repositories, including procurement routines and individual knowledge. Our second key concept was organizational design, which we define in structural terms based on the dimensions of structural differentiation and integration (Lawrence and Lorsch, 1967; March and Simon, 1958). Empirically, this entailed an examination of how tasks in the procurement process were divided among various organizational units and how these units coordinated their activities.

Below we provide a summary of the two studied cases focused on outlining the specific empirical setting, the timeline of the case and our observations concerning organizational structure and contractual learning. Figure 1 provides a dynamic overview of how organization, external feedback and post-signing contingencies relate to contractual design and learning in the two cases.

MunOrg is a municipality organized into departments, each responsible for a specific service area. Cleaning is a major activity, procured separately by each department with administrative support from the central procurement unit. Contracts typically cover part of a department’s operations and past two years, with a possible 24-month extension. Extensions depend on departmental satisfaction, which is informed by performance controls from the cleaning governance unit, cleaning support and supplier evaluations conducted by the procurement unit.

Organization: The procurement unit at MunOrg is responsible for procurement strategy, market analysis, contract development and evaluation and overall procurement administration. The most time-consuming activity is drafting contracts and discussing them with the reference group. This includes both economic considerations – such as market positioning and pricing models – and ensuring legal compliance. Since 2012, MunOrg has moved toward a more centralized, functionally specialized structure. As part of this transition, the procurement unit was expanded and relocated from the service department to the central municipal administration. At the time of the study, it employed around 30 staff across three sections: IT, services and consumables. This scale and specialization have enabled standardized processes tailored to specific procurement categories. The cleaning support unit manages supplier communication and quality control based on contractual requirements. Its primary responsibility is technical contract governance using the INSTA800 cleaning standard, which defines cleanliness levels and corresponding evaluation methods. Initially launched in 2012 as a temporary project, cleaning support later became a permanent unit, which in 2019 employs five cleaning coordinators. These coordinators, experts in technical aspects of cleaning, maintain close contact with both suppliers and department representatives, providing first-hand insights into service performance. This operational knowledge is then fed back into the development of new procurement contracts.

Contractual learning: An analysis of the contractual templates used for cleaning services shows that they have been subject to significant changes. Contractual changes involve more extensive supplier obligations (e.g. quality controls, employee competence/protection and damage liability) and more extensive control-oriented terms giving MunOrg means of enforcing contractual rights (e.g. price adjustments, rights to withhold payment, penalties and termination). These contractual changes have typically been made in an incremental fashion between individual procurements based on the experiences made in the previous procurement. According to a cleaning coordinator, new signals about what MunOrg should change in their contracts are continuously being received by cleaning support. These signals may originate from the department where the cleaning is performed, from the supplier or from the daily follow-up and coordination work carried out by cleaning support. However, the task of revising contractual templates has been sluggish, and it was only after the unit brought in an external lawyer that the templates were updated on a more general level. An important event that provided external feedback and prompted contractual changes was a legal dispute with a cleaning supplier that MunOrg terminated because of perceived low-quality cleaning. According to respondents, this specific court case showed the importance of contractual specifications concerning quality along with strong enforcement and control mechanisms.

ResOrg is a publicly owned research organization that depends on the design, procurement and installation of several advanced technical systems. The focal project started in 2013 following a request from the technical unit responsible for installing one of the systems. This led to an open public procurement procedure in 2014, which attracted only two bids because of the niche nature of the industry and the high engineering complexity. EngComp, a European engineering firm, was awarded the contract based on a combination of lower price and superior performance on the evaluation criteria. As the project progressed, coordination problems emerged, particularly delays in delivery and installation. These delays increasingly threatened the timeline for testing and commissioning the interconnected systems. Some of the complications stemmed from ResOrg’s own change orders. Although the contract entitled ResOrg to substantial liquidated damages for late delivery, concerns arose that EngComp was growing anxious about the project’s financial implications and had adopted a more rigid stance on cost-related issues. This shift strained collaboration and further impeded coordination. To resolve the deadlock and stabilize the project, the technical project head initiated a contractual renegotiation. The revised agreement extended the delivery deadline and waived ResOrg’s claim to liquidated damages, in exchange for releasing ResOrg from payment obligations related to change orders.

Organization: Like all other large procurement projects at ResOrg, the contract with EngComp was managed in close collaboration between the technical unit, responsible for system design, specification and construction and the procurement unit, responsible for the economic aspects of sourcing new components. The technical unit, tasked with managing the procured system, is relatively small, comprising about five specialist engineers and researchers. The procurement unit had been consistently growing over the timespan of the project, and at the time of the study, it consisted of approximately 13 employees. At the time of the original contract with EngComp, ResOrg’s legal function operated as an independent department. However, after a reorganization, which integrated the legal department into the procurement unit, legal services related to procurement and contracting were directly provided by two legal counsels embedded in the procurement unit. This reduced formal barriers, allowing for more seamless communication and collaboration between legal counsels and procurement officers. This new organization played a significant role during the contract renegotiation with EngComp, where procurement officers and legal counsels worked together to modify contract terms to address the project’s technical and economic contingencies effectively.

Contractual learning: The original EngComp contract was designed by the head of the legal department. According to several respondents, the design of the original contract with EngComp was largely a product of ResOrg being a young organization at the time without extensive experience in designing contracts for large technical projects. For example, ResOrg did at the time of the original contract not have a clearly developed procurement strategy to guide the design of the contract. This lack of experience and a clear procurement strategy affected ResOrg’s choice between types of procurement processes, the timing of the contract, how the responsibility for installation was regulated in the contract and how to regulate liquidated damages claims. According to ResOrg’s interpretation of the original contract, the liquidated damages related to delays were extensive at the time of the renegotiation. However, to avoid jeopardizing the relationship and the future delivery of the system, ResOrg chose to waive these claims in the renegotiated contract. Hence, key decision-makers at ResOrg changed priorities as they gained new experiences during the project concerning the uncertain contingencies associated with the execution of the contract, where the adaptiveness of the contract were given greater weight relative the legal penalty clauses included in the original contract.

The case studies of MunOrg and ResOrg showcase different forms of contractual changes that cannot be attributed to changes in transaction attributes. In the case of MunOrg, this included changes in contractual templates toward more extensive control-oriented terms related to penalties and termination. In the case of ResOrg, contractual learning involved a revision of liquidated damages clauses to facilitate technological adaptation and better relational dynamics. This section outlines the organizational factors that we found to be most important for facilitating these changes. A summary of key findings in the two cases is presented in Table 2.

Functional specialization and standardization played an important role for the development of new knowledge concerning how to design and govern contracts in both studied cases. ResOrg’s procurement organization was at the onset of the studied project based on three specialized functions or units: technical, economic and legal. While the procurement organization was underdeveloped and lacked the support of a clear procurement strategy in the early phases of the EngComp project, the clearly defined knowledge sets and professional roles contained in these three functions played an important role in facilitating the design and negotiation of the contract with EngComp. However, the case also illustrates how the articulation and codification of procurement knowledge and the different steps of the procurement process may enable contractual learning. According to respondents, these aspects were less developed at the time when the EngComp contract was entered, but gradually improved as the project progressed, first in the form of a procurement strategy that outlined the general direction of procurement activity at ResOrg and later in the form of a manual for contract management that standardized the contractual process.

Functional specialization in the case of MunOrg was initially less pronounced. In fact, addressing this problem was the major concern identified by the procurement manager when initiating the expansion and reorganization of the procurement unit. For cleaning services, this problem was addressed in three steps. The first step involved the centralization of procurement support across all departments to the procurement unit. The increased demand for procurement services allowed the unit to grow in terms of number of employees, which in turn created the conditions for specialization in terms of sections focusing on different types of goods and services and more rigorously codified routines and processes. The second step was the creation of a technically specialized contract governance function in the form of the cleaning support unit. Because of the broad procurement needs of a municipality, developing sufficient internal technical knowledge within each procurement area constitutes a significant challenge that may affect contract design, governance and the potential to learn from governance activities. With the cleaning support unit, MunOrg gained an opportunity to not only push increased technical specialization in cleaning, but more importantly, to also significantly improve contractual governance, which in turn strengthened the feedback loop between the governance and design of procurement contracts. The third step involved incorporating previous experiences into contractual templates to be used in new procurements, which was eventually finalized with the involvement of an external law firm. This last step points to a potential shortcoming in MunOrg’s legal specialization in the procurement process. The fact that MunOrg, over a period of several year, was unable to implement a major overhaul of its contractual templates show how straightforward tasks related revising contractual templates may require dedicated attention and specialist legal knowledge.

The organizational integration of the economic and technical functions in the procurement process was a key factor affecting contractual learning in both MunOrg and ResOrg. In MunOrg, the procurement process for cleaning services was built on a close collaboration between the procurement officers (economic function) and the cleaning coordinators (technical function). The procurement officers were responsible for market analysis, preparing procurement documents and tenders, whereas the cleaning coordinators were responsible for technical matters, such as the development of procedures and adoption of standards, quality controls and supplier communication. When a municipality department perceives a cleaning need, it will typically approach the procurement unit with this request. The procurement unit then asks cleaning support to visit the concerned premises and speak to department representatives to specify the cleaning need in an operational format. In this activity, the technical cleaning knowledge of the cleaning coordinators plays an important role in articulating the need of the department into specifications of the parties’ roles and responsibilities that are later picked up on by the procurement unit when designing the contract.

The second major task performed by cleaning support is post-contract governance in the form of visual quality controls performed according to contractual criteria. According to respondents, the key to effective contractual governance is the cleaning coordinators technical knowledge. This knowledge allows cleaning coordinators to perform controls in an accurate way and to transfer new insights about cleaning made in the controls to be included in future contracts. An important factor influencing MunOrg’s contractual learning, thus, involves the knowledge flows created when technical experts involved in the specification of department needs and hands-on governance activities are invited to weigh in on the evaluation and the design of new contracts. This illustrates how economic-technical integration may impact contractual learning in public procurement: First, cleaning support’s involvement in the early phases of the procurement process where buyer needs are specified provides an important input to learning related to coordination-oriented contractual terms (roles/responsibilities). Second, cleaning support’s involvement and joint work with the procurement unit in later contractual governance activities, such as quality controls, provide and important input to learning concerning control-oriented contractual terms (e.g. number of controls, penalties and termination clauses).

In the case of ResOrg, the procurement process starts with the responsible technical unit defining a need and creating a technical specification on a procurement request form, which is sent to the procurement unit who develops the procurement documentation, sets up procedures for tender and designs the contract. When the contract is in place, the responsibility for the project is handed back to the responsible technical unit who manages contract governance for the duration of the project. ResOrg, thus, retains a strong involvement of the technical function throughout projects, whereas the procurement unit is mainly involved in pre-signing activities related to the tender and contract design. Hence, technical-economic integration may change over the lifespan of a procurement project in ways that affect contractual learning concerning different terms and clauses. This is important for the type of knowledge that is developed because early or pre-signing phases of the procurement process are typically focused on developing technical specifications, whereas potential problems related to delivery, quality and project contingencies typically occur in the post-signing phases of the project. Hence, many of the issues that arise in a procurement project may never be directly experienced by the employees responsible for preparing tenders and designing contracts. As demonstrated by ResOrg’s EngComp project, such organizational barriers can impede contractual learning, particularly in terms of how contractual safeguards are aligned with technical terms and project contingencies.

A major concern when going into the EngComp project was ResOrg’s lack of a clear procurement strategy that specify how economic and technical project-specific concerns should be weighed against each other. According to respondents, one consequence of the lack of a clear procurement strategy that provided a basic template for economic-technical interaction was that the contract with EngComp did not properly balance technical concerns, such as project scope and the division of responsibilities between parties, against control-oriented terms, such as the use of penalties and liquidated damages. In other words, the control-oriented terms of the contract were not adaptive to the type of technical disturbances or contingencies that may be expected in the project (they were too restrictive). As explained by a procurement officer, this is one of the things that ResOrg addressed during the EngComp project by the development of a new procurement strategy, which regulated the organizational interface between technical units and the procurement unit:

Contractual learning differs from other types of organizational learning because it involves a significant legal aspect related to contract law and design. Legal matters are in many organizations handled by a specialized legal function (administrative lawyers, legal counsel, etc.). In the case of ResOrg, the legal function was initially (at the time of signing the EngComp contract) set up as an independent department. Respondents describe this organizational design as promoting the internal visibility and influence of the legal department, but at the same time leading to a lack of communication and disagreements between the legal and procurement unit. To address this problem, ResOrg initiated a reorganization that involved closing the legal unit and integrating legal counsels directly into the procurement unit. The organizational change, implemented after the signing of the original contract but before the contractual revision, significantly lowered barriers to involving legal in various procurement issues related to the EngComp project. According to a legal counsel, reducing economic-legal organizational barriers improved reciprocal information flows between legal counsels and procurement officers. With less social distance, more informal, project-specific issues were brought to the attention of legal counsels, enabling them to provide input without requiring a formal request between departments. Being part of the procurement unit allowed legal officers to access information that might not have been available through formal channels, thereby enhancing the quality of legal decisions. A respondent describes this change as follows:

In the EngComp project, the events leading from the initial contract to the contractual amendment illustrate how control-oriented terms in procurement contracts may be revised because of learning and how this process is affected by the interface between the legal and economic functions. The specific problem that had to be addressed in the EngComp contract was the relative weight placed on technical adaptiveness to uncertain contingencies versus the inclusion of contractual terms for liquidated damages. In the initial contract, the independent legal unit had pushed through extensive penalties in the case of delays or failure to deliver. When negotiating the contractual revision under the new organization where the legal function had been integrated into the procurement unit, these terms were jointly deemed unworkable from a technical coordination perspective. Hence, it was the joint realization by procurement officers and legal counsels working in the procurement unit, that the original contract was lacking sufficient flexibility for ex post contractual adaptations and that the project would benefit from an amendment that nullified several of the stricter control-oriented terms in favor of an approach more oriented toward technical coordination between the parties. As explained by a respondent:

Whereas procurement at ResOrg, both before and after the reorganization, involved a distinctive legal function that performed specified task related to contract design, MunOrg operated with a much less well-articulated differentiation between the legal and economic functions. As is common in municipalities, the procurement officers at MunOrg were expected to manage individual procurements and contracts without a direct involvement of legal specialists (the procurement unit had an administrative lawyer focused mainly on disputes and regulations). Given the background of many procurement officers, this led to a focus on the economic aspects of procurement contracts, while the unit struggled to develop the formal aspects of contractual templates and getting an overview of the contract design challenges it faced. Hence, while a comparison between cleaning contracts show that changes were primarily associated with making control-related terms stricter (e.g. increasing penalties and stricter decision-rights), these changes in the contractual templates had lagged significantly because of the lack of knowledge concerning how to redesign contractual templates. In fact, many of the needed changes were only dealt with in a structured way once an external lawyer was brought in to review and update the contractual templates. As explained by a respondent at MunOrg:

Previous research has not extensively studied how internal organization affect contractual learning, and there are few empirical studies of contractual learning in public procurement. This presents an opportunity to empirically contribute to previous research on contracting in public procurement (Girth and Lopez, 2019; Kleemann et al., 2012) and the internal organization of learning (Argyres and Mayer, 2007; Hallberg, 20024; Jansen et al., 2009).

In the cases we studied, contractual learning was most evident in the development of control-oriented terms (such as contractual safeguards and legal provisions). Both cases also demonstrate how this learning was influenced by organizational factors, including the consolidation of procurement activities into specialized units and the degree of structural integration between these units.

Functional specialization: The question of to what extent the organization of public procurement should be centralized or decentralized has been examined in the literature on public procurement (Glas, Schaupp and Essig, 2017; Kamann, 2007; McCue and Pitzer, 2000; Patrucco et al., 2019). In relation to this research, our empirical observations particularly highlight the importance of an organizational structure that supports contractual learning based on scale, functional differentiation and specialization. Hence, both studied organizations progressively increased the size of their procurement unit to allow for more specialized and standardized processes, which gradually improved contractual outcomes. These observations are consistent with previous research on contractual learning in the private sector, which suggest that contract management builds on specialized subfunctions performed by managers, engineers and lawyers (Argyres and Mayer, 2007). When organizations centralize procurement activity, they increase the frequency with which specialist activities may be performed and, thus, create an internal demand for specialized subfunctions within the organization. When managing procurement contracts (designing clauses concerning termination, quality assessment and liquidated damages), functional specialization is conducive to learning because the increased frequency with which specialized tasks are performed increases the intensity of experiential learning (Argote and Miron-Spektor, 2011), while also allowing the acquired knowledge to be more effectively codified and retained in standardized processes (Fiedler and Welpe, 2010). This explains why the consolidation and centralization of procurement activity, which is often accompanied by increased specialization and standardization, is likely to constitute an important means through which organizations can facilitate learning concerning how to design and govern public procurement contracts.

Economic-technical integration: Public procurement often involves complex tradeoffs between technical performance and economic considerations. These tradeoffs are further complicated by the fact that economic and technical expertise is in many cases dispersed across distinct units. As a result, effective procurement depends on integrating these specialized knowledge domains to coordinate complex tasks and ensure cohesive decision-making (Patrucco et al., 2019). Building on Lawrence and Lorsch’s (1967) concept of structural integration, we find that the degree of integration between technical and economic units plays a key role in shaping contractual learning. Integration facilitates the transfer and alignment of knowledge across units, which is critical when tasks are interdependent and require joint problem-solving (Tushman and Nadler, 1978). Prior research identifies several mechanisms for achieving integration, ranging from multifunctional teams and integrative units to more informal mechanisms, such as social cohesion and connectedness between employees (Jansen et al., 2009).

In the context of contracting, studies have shown that contractual learning tends to be more pronounced concerning the transaction-specific, technical aspects of coordination-oriented terms than for control-oriented terms (Vanneste and Puranam, 2010). Further, both technical and economic functions hold important and complementary sets of knowledge when developing coordination-oriented clauses (Argyres and Mayer, 2007). This has prompted researchers conclude that economic-technical integration specifically facilitates learning concerning coordination-oriented terms (Hallberg, 2024). However, somewhat surprisingly, our findings suggest that economic-technical integration also plays a critical role for contractual learning concerning control-oriented terms because of the strengthened feedback loop between the design and governance of contracts. In MunOrg, for example, the technical unit responsible for supplier monitoring provided feedback to the procurement unit based on its quality controls. The procurement unit then implemented progressive elaboration of penalty clauses and termination conditions in subsequent contracts based on this external feedback. This mechanism illustrates how integration between the governance and design phases can enable an organization to better anticipate and address supplier opportunism. In other words, stronger integration between economic and technical functions facilitates more effective learning concerning control-oriented terms by enabling more detailed experiential feedback from contract governance to contract design. More formally, we propose the following:

Economic-legal integration: Contractual learning differs from many other types of organizational learning because it involves an important legal aspect (Weber and Mayer, 2011). Legal issues are in many organizations the responsibility of a specialized legal function consisting of lawyers and legal counsels (Macaulay, 1963). One important challenge in facilitating contractual learning in public procurement is, thus, related to the involvement of legal specialists in the process. A legally astute organization is characterized by being able to use its legal function as a proactive agent in economic and technical decisions (Bagley, 2008). However, hiring and involving legal specialists in the procurement process requires scale, the consolidation of procurement activity and economies of specialization and is also likely associated with integrative barriers connected to the highly specialized nature of legal work (Nelson and Nielsen, 2000).

Our empirical observations show how legal integration in the procurement process may facilitate contractual learning by allowing the organization to align economic aspects of contracts with the need for contractual safeguards and legal governance mechanisms. In line with previous research (Hallberg, 2024; Weber and Mayer, 2011), we find that control-oriented contractual terms are to a higher extent dependent on these tradeoffs and, hence, more affected by the level of legal integration. However, and perhaps somewhat surprisingly, greater legal integration does not necessarily imply the development of stricter and more control-oriented contracts. In case of ResOrg, for example, greater legal integration improved the organization’s capacity for tailoring contractual safeguards (liquidated damages) in response to new technical and economic contingencies (delays) and external feedback (supplier dissatisfaction) by making penalty clauses less restrictive and more partner friendly. Comparing this development with the case of MunOrg where contracts were made more control-oriented following the external feedback of lawyers provides an interesting benchmark concerning the relationship between legal integration and contractual learning. Overall, the patterns observed in the studied cases, thus, indicate that the impact of legal integration on contractual learning in public procurement is primarily related to the organization’s ability to apply control-oriented terms in an adaptive and context-specific way. Hence, more formally, we propose the following:

This study is based on two case studies of public organizations, which naturally limits the generalizability of the findings. As with all case study research, there is a trade-off between breadth and depth: while the inclusion of additional cases might have yielded further insights, it would also have increased analytical complexity and likely reduced the level of detail attainable in each case. The selection of the two cases was aimed at maximizing variation along theoretically relevant dimensions, such as transaction frequency (routine service procurement versus one-off project) and uncertainty (standardized service versus customized, technologically advanced equipment). This design choice enabled a focused analysis of how public organizations learn to contract, while maintaining a high degree of within-case richness (Yin, 2018). Future research could extend these insights by examining additional types of public organizations, a broader range of goods and services and diverse national or regulatory environments to assess the robustness of the observed patterns.

A further limitation lies in our reliance on retrospective interviews and documentation, which may not fully capture the dynamics of learning as they unfold in real time. Prospective or longitudinal research designs could, therefore, yield deeper insights into the microfoundations of contractual learning in public procurement. In addition, while this study concentrated on internal organizational drivers of learning – such as the integration of legal, technical and commercial expertise – we did not systematically explore how transaction characteristics or relational dynamics influence learning processes. Future research might incorporate dyadic data collection from both buyers and suppliers, thereby illuminating how interorganizational relationships and market conditions shape contractual adaptation over time.

Finally, our findings indicate that formal safeguards and control-oriented clauses are particularly prevalent in public procurement because of legal constraints on contractual changes and terminations. Further inquiry into how such constraints affect the direction and pace of learning could contribute significantly to research on both public procurement and contractual learning. Our findings underscore a productive tension between TCE – which emphasizes contractual governance through legalistic safeguards in response to transaction attributes (Williamson, 1985) – and insights from capability-based theories, which highlight the role of organizational knowledge and structure for adaptive responses under uncertainty (Zollo and Winter, 2002). Hence, we suggest that future research on TCE examine whether, and under what conditions, organizational capabilities moderate the relationship between transaction attributes (such as asset specificity and uncertainty) and the choice of governance structure.

Respondents at both MunOrg and ResOrg emphasized that public procurement law constrains contractual adaptiveness and complicates the tradeoffs that practitioners must navigate. This view aligns with our initial premise that regulatory hurdles in the public sector limit the contractual options available and, thus, shape how organizations learn to design and govern procurement contracts. Our cases highlight two organizational solutions that appear effective.

Dedicated technical units for contract governance: In the case of MunOrg, the setup of the cleaning support unit was generally upheld as a major success by respondents that should be used on a broader scale beyond just cleaning services. Installing a specialized technical unit responsible for the contractual governance of a particular type of service, such as cleaning, had the benefit of focusing specialized technical knowledge in the post-signing phase of the procurement process, which created a strong feedback mechanism that played an important role for finding new solutions and developing better procurement contracts. Yet such specialized technical units appear to remain rare in public organizations, perhaps because of the broad range of goods and services they procure.

Integration of legal competence: In the case of ResOrg, the decision to integrate the formerly independent legal unit into the procurement unit was upheld as an organizational advance that enhanced feedback and information flow between procurement officers and legal counsels. Respondents particularly highlight the important role of informal discussion and feedback, which may be hampered by formal inter-unit processes and departmental procedures that may invoke feelings of prestige. The case of MunOrg, by contrast, offers an example of the negative effects on contractual learning that can arise from a lack of legal involvement, leading to delays in the revision of contractual templates.

In this paper, we have addressed the question of how organization influence contractual learning in public procurement. Based on two case studies of procuring organizations, we present empirically grounded propositions concerning the effect of organizational structure on contractual learning in public organizations. Our results highlight the importance of specialized technical, economic and legal knowledge in the procurement process; as well as the differential impact of integration mechanisms on the type of learning that is likely to occur.