The impact of the blockchain on the supply chain: a theory-based research framework and a call for action

2.1 Industry adoption of the blockchain

It was Bitcoin, a cryptocurrency application of the blockchain, which first drew public attention to this technology. In 2008, a seminal paper entitled “Bitcoin: A Peer-to-Peer Electronic Cash System” was published under the name of Satoshi Nakamoto, (Nakamoto, 2008), and in January 2009, the first open source Bitcoin client was released with Nakamoto, himself being the first miner. This name was later revealed to be a pseudonym and the actual inventor (or group of inventors) of Bitcoin still remains unknown. In the years to follow, it was the computer science and cryptography communities that were most occupied with the technology, but attention broadened around 2010/2011 when Bitcoin started to take off with exponential trade volumes and a dramatic price increase (Alabi, 2017).

Bitcoin has paved the way for a multitude of so-called Altcoins to follow, many of which were actually derived (“forked”) from the original Bitcoin implementation. Those Altcoins serve different purposes ranging from replacing fiat money to more specific applications such as supplanting the use of the USA dollars in the container shipping industry with tokens launched by open-source blockchain-based platforms (MarEx, 2017). Current use cases go far beyond simply using the blockchain as a technological platform for currencies and have now started to include the implementation of so-called smart contracts, for example, on the Ethereum platform. This was made possible by an extension of the basic functionality of the blockchain that enabled the implementation of complex business logic. Smart contracts are agreements between two or more contracting parties which can be automatically enforced without any intermediaries. These agreements exist in the form of software code on the blockchain platform, which ensures their autonomy and self-executive nature based on predefined rules (Savelyev, 2017). The blockchain is frequently labeled a “disruptive” technology, but, as Iansiti and Lakhani (2017) point out, it would be more appropriate to call it a foundational technology, as it can create new foundations for social and economic systems.

Although at present the full range of potential blockchain applications is still unclear, many use cases have been suggested for various kinds of industries. A notable example in manufacturing, SCM and logistics involves combining the internet of things (IoT) and the blockchain. This combination has the potential, for example, to enable connected vehicles that automatically communicate their current status and smart locks that give access to real-world objects. Blockchain platforms can also be used for SC tracking applications which keep a formal product registry and are able to track possession of goods at different stages in the chain. Other ideas include smart diagnostics, which will enable machines to monitor their state and diagnose problems and smart contracts between manufacturers and vendors for the procurement of supplies and service of machinery (Bahga and Madisetti, 2016). Additionally, it would be possible to track the identity and reputation of suppliers and to use smart contracts to automatically negotiate best prices in real time while taking into account the seller’s reputation (Bahga and Madisetti, 2017). On a more general level, Field (2017) predicts that the blockchain will lead to improved SC visibility, SC optimization and better demand forecasting, as increased transparency will make it possible to react in real time to unforeseen events. This will also have a positive effect on inventory management. The general expectation is that the blockchain will help to reduce fraud and errors, reduce transit and shipping costs as well as waste and improve inventory management (JOC, 2017).

4.1 Complementary theories in SCM/logistics management

Agency costs within PAT can be defined as the sum of monitoring expenses by the principal, bonding expenditures by the agent and residual loss (Jensen and Meckling, 1976). They occur as the principal wants to monitor, supervise and control the agent, so that the latter acts in the best interest of the former. The principal has incomplete information regarding the agent’s behavior and therefore has to trust the agent to a certain extent. Fayezi et al. (2012) identified 86 SCM-related articles dealing with agency theory and principal–agent relationships, 19 of which they closely investigated. Simply put, the major challenge for the principal lies in selecting the right agent and building a relationship based on trust. Previous principal-agent studies related to SCM have focused on issues as diverse as the impact of Sarbanes–Oxley on off-balance sheet SC activities (Kros and Nadler, 2010), performance improvements (Miller et al., 2013), demand uncertainty and price competition (Narayanan et al., 2005), SC quality management (Zu and Kaynak, 2012), the likelihood of quality fade in buyer-supplier relationships (Whipple and Roh, 2010) and the reduction of supply risk (Zsidisin and Ellram, 2003).

TCA postulates that the selection of the most efficient types of inter- and intra-organizational structures impacts the efficiency of different types of governance mechanisms in terms of production and transaction costs (Coase, 1937). Transaction costs include ex ante costs of initiation (search and information) and agreement (negotiation and decision-making) and ex post costs of control and adjustment. Depending on the costs of market transactions, different governance structures, such as markets or hierarchies, might emerge (Williamson, 1987; Picot et al., 1997; Strebinger and Treiblmaier, 2006). TCA provides a rich theoretical base for SCM, and measurement tools, guidelines and research propositions have been developed to investigate topics such as outsourcing vs purchase decisions, allocation of investments and SC coordination, integration and distribution (Grover and Malhotra, 2003). Exemplary cases in SCM include the application of TCA in the context of information sharing (Yigitbasioglu, 2010), suppliers’ adoption of environmental practices (Tate et al., 2011), purchasing portfolio management (Luzzini et al., 2012), performance effectiveness of relational and contractual governance (Wacker et al., 2016) and the perception of transaction risks (Stranieri et al., 2017).

The RBV emerged as a countermovement to the positioning school, which emphasized the importance of strategy for a particular organization (Porter, 1980). In contrast, the RBV postulates that it is only a subset of a company’s resources that creates competitive advantage and an even smaller subset which leads to superior performance in the long term. The possession of scarce resources leads to the creation of competitive advantage which can be sustained for as long as the company can protect its resources and no substitutes emerge (Wernerfelt, 1984; Barney, 1991). Halldórsson et al. (2007, p. 288) stated that “only a few articles have applied the resource-based view (RBV) to the field”, but a growing number of studies have highlighted the usefulness of this theory and closely related concepts such as the resource-advantage theory (Hunt and Davis, 2008) for SCM research. Amongst others, more recent SCM articles have investigated the impact of companies’ resources and capabilities on logistics performance (Yang and Lirn, 2017), the role of market information in creating an informational advantage for transportation providers (Golicic et al., 2012), interorganizational learning (Manuj et al., 2013), achievement of closed-loop SC designs (Miemczyk et al., 2016) as well as the determination of SC information integration antecedents (Huo et al., 2016).

NT seeks an understanding of the dynamics of interorganizational relations by focusing on personal relationships between the parties as well as the mutual creation of trust through cooperative relations and exchange processes (Halldórsson et al., 2007). To gain access to external resources, companies have to establish relationships. On a larger scale, this creates networks that are both stable and changing. Two interaction types, exchange processes (social, business and informational) and adaption processes (products, production and routines), help to establish stable links within a network of firms (Johanson and Mattsson, 1987). Important aspects of NT have been incorporated in a wide variety of SCM topics, including, for example, manufacturing joint venture formations (Carnovale and Yeniyurt, 2014), network centrality antecedents in an environmental SCM initiative (Wichmann et al., 2015), the functioning of strategic networks and alliances (Klint and Sjöberg, 2003) and networked versus non-networked companies in the software industry (Kulmala and Uusi‐Rauva, 2005).

Table I contrasts the major characteristics of the four theories. All of them assume a certain level of bounded rationality but take a different problem orientation when it comes to investigating the underlying object of interest. PAT is mostly focused on how to design optimal contracts to avoid problems of asymmetric information. TCA explores the borders of organizations resulting in an optimal organizational size and efficient governance structures. RBV focuses on differences between companies and investigates the internal competences that help companies to gain a competitive advantage and, finally, NT deals with dyadic relationships and the networks in which they are imbedded (Halldórsson et al., 2015). Accordingly, Table I also lists the key questions one needs to ask when applying the respective theory, the primary focus of analysis, the nature of relations and the primary domain of interest. All four theories thus take a different stance when it comes to explaining and exploring real-world SCM phenomena; combining them for methodological plurality allows for a more comprehensive investigation of a multifaceted research topic than would be possible with a single theory (Georgi et al., 2010).

4.2 A framework for blockchain research in SCM/logistics

Analyzing the potential implications of the blockchain from different theoretical angles provides promising starting points for further theory development and empirical research. The main characteristics of the blockchain include decentralization of decision making, distributed processing, reliability, peer-to-peer transmission, immutability of data, automaticity, speed, low transaction fees, transparency with pseudonymity and non-repudiability (irreversibility) (Nakamoto, 2008; Narayanan et al., 2016; Bahga and Madisetti, 2017; Iansiti and Lakhani, 2017). These characteristics lead to derived concepts on a higher level which bear substantial managerial implications such as provenance of data, trust, privacy, security, enforcement, integrity, consensus, availability, authenticity and accountability (Tapscott and Tapscott, 2016; Tschorsch and Scheuermann, 2016; Liang et al., 2017; Neisse et al., 2017) all of which might have severe implications for the SCM/logistics field (Field, 2017; Hackius and Petersen, 2017; JOC, 2017; MarEx, 2017).

Table II shows how these attributes of the blockchain can be integrated into the four theories by adapting the framework from Table I. The structure of this table (i.e. the respective categories) was taken from Halldórsson et al. (2007) who applied this framework to third party logistics. The topics shown in Table II were gained by applying the characteristics of the four theories, as shown in Table I, to blockchain technology. This process was done by the authors of this paper based on the existing literature and on independent feedback from three additional researchers. The characteristics of the four theories were sent via e-mail together with an invitation to propose the most important topics. After the initial round the results were sent to all researchers for final confirmation.

The most important findings were that, from a PAT perspective, smart contracts have the potential to solve or alleviate the problem of information asymmetry (ISDA, 2017). TCA can help to explain how the costs of intra- and inter-organizational transactions will change through the blockchain and to what extent this will impact organizational borders (Cocco et al., 2017). RBV can help with the allocation of resources to gain competitive advantage. The specific characteristics of the blockchain are expected to render certain advantages obsolete but might also provide strategic opportunities for fast adopters (IBM, 2017). Finally, the blockchain’s promise to create “trustless trust” (Werbach, 2018) can be systematically analyzed by applying NT.

As the logistics field has evolved from a rather description-based discipline to one that is founded upon solid theoretical grounding, a need for middle-range theorizing has arisen (Stank et al., 2017; Craighead et al., 2016). As opposed to general theorizing, middle-range theories are more specific regarding the field, of analysis and focus on a better understanding of why, how and when specific outcomes occur (Astbury and Leeuw, 2010). A complementary taxonomy of theories classifies them according to their methodological approach, which can be analysis, explanation, prediction, explanation plus prediction or prescription (Gregor, 2006). While analysis is mainly a descriptive approach focusing on “What?” questions, explanation answers “Why?”, “How?”, “When?” and “Where?” questions. Theories focusing on prediction create testable propositions but may lack justifiable causal explanations, while design and action research gives prescriptions for constructing artifacts (Gregor, 2006).

Table III lists the different methodological starting points and the respective research objects together with those key questions that can be used to derive testable and theory-based propositions, hypotheses and models. To better understand the potential implications of the blockchain on the respective research objects, it is useful to thoroughly analyze changes in relations between principals and agents, organizational boundaries, critical resources and relationship structures. In contrast, a predictive research design will focus on the changes that are to be expected and a design and action approach makes prescriptions on how to deploy blockchain applications to achieve desired results. As far as middle-range theories are concerned, explanatory questions are of the greatest significance, as they focus on the underlying rationale of why, how and under which conditions certain phenomena occur. Applying these three core explanatory questions to the respective research objects within a SCM/logistics context will help to create middle-range theories that are based on substantiated previous research but at the same time are specific enough to provide insights that are relevant for the logistics industry.

Formulating middle-range theories starts with accumulated knowledge about a phenomenon within a particular domain. This knowledge may stem not only from previous theory-based frameworks but also from inductive, qualitative observations of practice (Stank et al., 2017; Kim et al., 2009). Combining the respective research objects and the key questions listed in Table III enables the deduction of research questions that are directly related to the implications of the blockchain on the SC. In the following paragraphs, we will briefly describe and visualize the expected changes from the points of view of the four theories, respectively, and present exemplary research questions that are of importance for SCM/logistics research.

The central feature of the PAT is the complex relationship between principal and agent, which is characterized by the asymmetric information flows depicted in the left part of Figure 1. It is only the agent who receives information from various external sources (e.g. customers, suppliers and partners), processes this information and acts upon it. As a consequence, the principal has to trust the hired agent and can only assess their general performance but lacks knowledge about details. Under the assumption that the agent will naturally act according to self-interest, the principal has to trust that the agent will pursue the principal’s interests, which may require the establishment of trust mechanisms and control systems at additional cost (Jensen and Meckling, 1976). In a blockchain-based scenario (right part of Figure 1), however, information flows are transparent and accessible to both the principal and the agent. As a consequence, the demand for trust vanishes. Blockchain technology thus provides formal guarantees to both parties that comprehensively address agency problems (Kaal, 2017).

Table IV lists some central research questions that arise from applying the questions in Table III to situations in which the blockchain is introduced, as predicted by scholarly as well as practitioner literature. All of the respective questions can be embedded into the general setting of PAT but provide enough specificity to constitute a middle-range theory that leads to useful outcomes for SCM research. All explanatory questions starting with “Why” shown in the following tables are based on the assumption that an effect exists, which must first be confirmed in a preliminary study.

TCA focuses on transactions costs that determine the size and structure of organizations. The immutability of blockchain records in combination with cryptographic security creates new forms of trust that are based on transparency and traceability. The rules that are embedded in blockchain code cannot be circumvented without detection (Kaal, 2017). In the case of smart contracts, the rules are enforced automatically without any need for human intervention.

Figure 2 shows different types of transaction costs and how they can be affected by the blockchain. Transaction costs can be internal (only shown for company A in Figure 2), representing bureaucratic costs that occur inside a company or external and relating to exchanges between companies. On the left side of Figure 2 company A conducts business with two other companies, B and C. The adoption of blockchain technology potentially affects internal and external transaction costs simultaneously and thus can lead to the shrinking or expansion of a company’s boundaries (indicated by the dotted boundaries and the arrows on the right side of Figure 2). Additionally, from the point of view of company A, it might no longer be necessary to continue business relations with companies whose services can be substituted by the blockchain (i.e. company C). This is predicted, for example, for services as diverse as clearing, trade settlement and regulatory reporting (Harwood-Jones, 2016). On the other hand, new market players offering services needed for blockchain technology might emerge (company D). In a nutshell, all internal and external transaction costs may be affected in nature and size. To avoid clutter, disintermediation is not shown in Figure 2, but it can easily be conceived that services from companies which currently benefit from being information and service hubs might become obsolete. In the transportation and logistics field this includes various services related to the validation of transactions.

Table V shows research questions that arise when TCA and blockchain are combined with the goal of creating middle-range theories.

By applying RBV those resources can be identified and analyzed that help a company to gain and sustain competitive advantage. Figure 3 differentiates between resources in general, capabilities that distinguish a company and finally, those resources which enable competitive advantage on the market. blockchain technology might lead to transformation processes that affect all three levels simultaneously, as is indicated by the dotted lines (right part of Figure 3). This can either be an increase or a decrease of the respective resource, capability or competitive advantage. Smart contracts, for example, have the potential to automate processes by running rules encoded in computer programs. Agents can be hired by so-called decentralized autonomous organizations (DAO), which are run by smart contracts, to perform specialized tasks (Morabito, 2017).

As is shown by the dotted lines representing new boundaries in Figure 3, such a scenario can lead to a reduction of unique competencies for those companies that offer services which can be potentially substituted by the blockchain, but it can also lead to an increase in competencies for those companies that can use the new technology to complement their own portfolio of services being offered. On the lowest level, attributes of the blockchain such as data transparency and non-repudiability offer opportunities for cost reduction and independence from the services of intermediaries. On the highest level, new business models may emerge and existing companies might find innovative ways to incorporate blockchain technology into their portfolio of services. The blockchain has been claimed to hold the potential to “disrupt the disruptors” and companies such as Uber and Airbnb, which build their competitive advantage on centralized databases and control structures, might no longer be able to sustain their existing business models (Ogundeji, 2016). Similar to the internet, blockchain technology thus bears the potential to fundamentally alter the importance of (critical) resources for sustained advantage across industries independent of factors such as company size or managerial experience.

Table VI presents several research questions that investigate the impact of the blockchain from an RBV perspective and which can serve as guidelines for middle-range theory building.

NT investigates the nature and role of interorganizational relationships and how they can be managed. NT is similar to TCA in that it investigates connections between organizations but focuses on the management of relations rather than transactions (Rinehart et al., 2004). On the left side of Figure 4 two types of relationships are shown: a direct one between A and B and an indirect (mediated) one between A and D. On the right side, the potential impact of the blockchain is shown. The relationship between companies A and B still exists, but its nature and strength might have changed. Additionally, the previously indirect relationship between A and D was substituted by a direct one and the intermediary C was removed. Then again, a new company E enters the market and serves as an intermediary between A and B, which also have a direct relationship.

Figure 4 depicts just some of the changes that can be induced by the blockchain in complex networks of interorganizational relationships. It is again the transparency and non-repudiability of the data which might render personal contacts for enhancing trust obsolete. This is not to say that interorganizational relationships become unnecessary in general, but their overall importance and qualities might change as they are no longer needed to ensure trust in SCs (Werbach, 2018). Similarly, the automated “behavior” of smart contracts makes the execution of business processes predictable. Blockchain technology therefore has the potential to significantly alter the importance of interorganizational relationships and to enable trusted information flows between hitherto disconnected companies, which in turn poses new challenges for management.

In Table VII exemplary research questions are listed that explore the impact of the blockchain from a NT perspective. Again, these questions refine NT such that middle-range theorizing becomes possible.

6.1 Theoretical implications

One of the major goals of any scientific discipline is to create and advance theory. Although original theory building efforts in SCM/logistics are scarce (Touboulic and Walker, 2015), the refinement and adaptation of existing grand theories into so-called middle-range theories bears a huge potential to closely investigate the conditions under which logistics phenomena yield certain outcomes (Stank et al., 2017). This paper presents a starting point in the development of middle-range theories by using four well-established and complementary theories (PAT, TCA, RBV, NT) and applying them to postulated changes induced by the blockchain: a technology that has the potential to restructure many logistics/SCM processes (Hackius and Petersen, 2017). The four theories cover both structural (PAT, TCA) as well as managerial (RBV, NT) aspects and, taken together, provide a comprehensive framework for designing future SCM research. Applying the characteristics of the respective theories to the specifics of the blockchain helps researchers to identify potential research areas (cf. Table I).

Following guidelines from the methodological literature (Gregor, 2006), different types of theories are distinguished according to their respective goals (analysis, explanation, prediction, design and action) and research questions of each theory type are identified for each of the theories, as presented in Tables IV-VII. These questions should guide researchers in the further pursuit of relevant and theory-based research topics. More specifically, each of these questions can serve as a starting point for research projects with the goal of analyzing, explaining, predicting and designing the impact of the blockchain on SCM from a specific theoretical angle. Our suggested framework does not impose any restrictions pertaining to methodology and is therefore suitable for researchers with both qualitative and quantitative backgrounds.

Furthermore, basing research questions on existing theory allows for the incremental development of a body of research that makes it easier to replicate and, if needed, to dismiss previous findings. Refining theory in this context not only pertains to investigating (i.e. explaining, exploring or predicting) the blockchain as a research phenomenon but also includes research in the “design and action” category that aims at designing and implementing concrete industry solutions. Incremental academic research in this area is especially desirable, as industry projects are often conducted behind closed doors and gains in experience are rarely shared. Academia can help in this regard, for instance, by highlighting best practices for the industry.

6.2 Managerial implications

The major part of the existing literature discussing the blockchain is practitioner-oriented and predicts a huge impact of the technology. This is also reflected in current investment levels and industry expectations (Holden and Moar, 2017). Only a few authors warn against raising unrealistic expectations (Walker, 2017; Coppola, 2016). In spite of the huge total investments by the industry, it is mainly the big companies who are at the forefront of this innovation and many logistics operators, especially small and medium-sized companies, lack knowledge about the blockchain (Kersten et al., 2017). To critically and impartially assess its impact, including its potentials and limitations, a repository of methods, frameworks and models are needed that can provide interesting insights for the industry. Academically sound research results that are based on empirical findings and well-founded reasoning – rather than on speculation – will benefit decision makers at all levels including commercial organizations as well as governmental and nongovernmental organizations.

In this respect, the logistics/SCM community can build upon a long tradition of theory-based research by adapting existing theories to problems that are of interest for the industry. The four theories that serve as a starting point in this paper are general theories requiring further refinement to create middle-range theories that lend themselves to validation and testing. In this paper, we show how this can be achieved by deriving theory-based research questions that are also relevant for the industry. Combining PAT, TCA, RBV and NT with four theory types according to their respective goals (i.e. analysis, explanation, prediction, design and action) leads to research questions that can help to better understand the implications of the blockchain, as well as providing indications on what is needed to actively take part in the design process.

When it comes to structural issues, PAT explains how complex inter- and intra-organizational dependencies might vanish because of increased information transparency. This guides companies, for example, in redesigning the structure of existing systems (e.g. the level of (dis)intermediation) and critically assessing the information requirements for a specific SC designs. In the next step, such an assessment can be used to design SCs that are more resilient by applying an optimized level of (de)centralization (Treiblmaier, 2018). From a TCA perspective, the focus is on changes in transaction costs and subsequently, in the structure and size of organizations. By investigating the research questions proposed in this paper managers will better understand how attributes of the blockchain can lead to adjustments in organizational size and the design of contractual agreements. This especially pertains to the impact of smart contracts, which are automatically executed and may severely impact transaction costs (Savelyev, 2017).

If the focus is on managerial issues, RBV can be used to critically assess the role, importance and growth potential of existing organizational resources. Managers might deduce, for example, that a strategic reorganization and delegation of responsibilities is needed to cope with blockchain-induced changes. Such changes may include positive impacts (e.g. an increase in competitive power because of new payment channels) as well as negative impacts (e.g. new competitors that offer simplified processing of shipping documents and help to track goods) (Kshetri, 2018). Finally, NT provides the underlying rationale for assessing the role and importance of companies’ interorganizational networks from a relational perspective. For example, analyzing the interplay between relationships and information transparency might inform managers on whether personal relationships can indeed be substituted by the increased information control offered by the blockchain. Additionally, the ease with which new business relationships can be established might change substantially because of “trustless systems” that enforce contractual compliance without any need for personal relationships (Kiviat, 2015).

Taken together, the logistics/SCM community possesses the tools and the methodological knowledge needed to tackle the research questions derived and presented in this paper in a systematic manner. The results from these studies will benefit managers who seek an unbiased, well-founded and thorough understanding of how blockchain technology is impacting and can further impact their own companies and SC networks.