Back to origins: Rediscovering the role of individuals in the Uppsala model

Valdemarin, Stefano

doi:https://doi.org/10.59876/a-2zh4-c5r2

The Uppsala model is one of the most prominent models in International Business (IB). Created by Johanson and Vahlne more than 45 years ago to describe the international development of companies, it evolved over time moving its focus from internationalization to evolution in networks (Vahlne & Johanson, 2017). Today, the Uppsala model faces a new challenge: integrating the individual dimension into the evolution process (Coviello et al., 2017; Vahlne & Johanson, 2020). This dimension, here labeled as the micro-level, focuses on individuals, viewing them as key actors of meso-level (i.e., organizational) processes.

A growing number of scholars point out that the micro-level is considered as a “black box” (Vahlne & Johanson, 2017, p. 1087) in the Uppsala model, and are calling for a multilevel approach to enhance the model’s explanatory power. Galkina & Chetty (2015, p. 671) propose “that future studies on entrepreneurial networking during internationalization could concentrate on individual entrepreneurs as the unit of analysis. […] entrepreneur’s networks at the individual level often overlap with the networks of their ventures”. Maitland & Sammartino (2015, p. 733) point out that “despite the cognitive foundations of several key constructs, standard internationalization models do not explicitly incorporate managerial cognition”. Similarly, Coviello et al. (2017, p. 1160), referring to the Uppsala model, propose that “future research might investigate how individual-level characteristics impact state and change variables of the MBE [multinational business enterprise]”. Moreover, it is suggested that “reflecting on micro-level assumptions can enhance actionable managerial implications of IB theory” (Kano & Verbeke, 2019, p. 117), particularly because, according to microfoundations theory (Contractor et al., 2019), individuals play a key role in the evolution of companies and consequently their international development.

The failure to include the micro-dimension can seriously harm the evolutionary nature of the model itself and hence its explanatory power. The model is based on a deep interplay between state and change variables. The latter constitute the dynamic engine “where the action takes place” (Vahlne & Johanson, 2017, p. 12) and are based on individual processes. As Coviello et al. (2017, p. 1156) explain “the lack of focus on individual relationships […] makes their evolutionary model of the MBE static. […] Vahlne and Johanson (2017) do not address the individual in their theorizing. As a result, we lack an understanding of the mechanism of change”. State variables, on the other hand, measure the level of commitment/performance and capabilities, but currently take no account of the individual. However, it is commonly acknowledged that “individuals […] greatly affect the behaviour, evolution, and performance of organizations […]. From this perspective individuals in organizations serve as microfoundations of routines and capabilities in various ways” (Felin et al., 2012, p. 1358).

The authors of the Uppsala model note the “urgent need for studying the interdependency between the nano- and micro-levels of internationalization. We believe this will be a core area in future IB research” (Vahlne & Schweizer, 2022, p. 1553). Moreover, the adoption of a multilevel approach will help the model to respond to a number of long-standing criticisms: “We believe that a focus on individuals’ behavior during the internationalization process is necessary to understand that the transition mechanisms of the Uppsala model indeed also incorporate discontinuous and non-linear internationalization patterns” (Schweizer & Vahlne, 2022, p. 586). In other words, placing greater emphasis on the role of individuals might shift the focus from the speed of internationalization to its deepest mechanisms.

After more than 45 years of evolution, (re)introducing individuals into the model is not an easy process. Our paper aims to rediscover the foundations of the Uppsala evolution model and to discuss the potential for interaction between the micro- and meso- levels of analysis on the one hand and the role of individuals on the other by looking at how the model has developed over time. Our research question can then be formulated as follows. Is it possible for the Uppsala network evolution model to become a multilevel model? This question includes two sub-questions. How does the model currently integrate individuals in its framework? And what theoretical concepts can the Uppsala model use to develop a multilevel approach?

Many studies and reviews of the Uppsala model have been conducted in the past from different angles, and a number of special issues have been dedicated to it (e.g., Meier et al., 2010). These studies variously carry out in-depth analyses of specific concepts mobilized by the model (Forsgren, 2002); look at the internationalization path and analyze entry modes (Cheriet, 2015; Meier et al., 2010); compare progressive and rapid internationalization (Knight, 1997; Oviatt & McDougall, 1997; Tapia Moore & Meschi, 2010); assess the validity of the model for specific contexts such as emerging markets (Angué & Mayrhofer, 2010); compare it with alternative approaches (Meier & Meschi, 2010); and illustrate how the model has shifted its focus from internationalization to evolution (Vahlne, 2020). However, none of these reviews address the role of individuals and how they can be integrated into the Uppsala model.

In a context where the authors themselves are discussing the role individuals can play in the model and are attempting to re-engineer their iconic framework into a multilevel configuration, our paper aims to contribute to the ongoing debate not only by conducting an exhaustive review of recent developments, but also by exploring the foundations and theoretical building blocks of the model. Along with its methodological outcomes, our review makes several contributions to the existing literature. It checks the possibility of introducing individuals into the Uppsala model in accordance with its theoretical underpinnings; explains how and to what degree the micro level can enter the existing framework; identifies which individuals are currently addressed by the model both in internal and external networks; proposes to build a more layered ontology by taking into account intermediate entities and sub-units; and focuses on the characteristics of individuals, including the way they develop networks.

Methodology

In order to answer our research question, we developed an in-depth, systematic literature review (Petticrew & Roberts, 2008) on the concept of individual networks within the Uppsala model. Although systematic reviews “may not be appropriate when reviewing a broad topic” (Fan et al., 2022, p. 173), a critical approach of this kind is used to highlight the importance of a specific, sometimes niche or emerging topic within an established framework or model (e.g., Forsgren, 2002). This is the case for our study, which is not oriented to look at the Uppsala model in general, but instead focuses on the role of individuals within it. The review process is summarized in Figure 1.

Figure 1

**The literature review process**

A succession of steps was taken to identify the publications to be included in our review. First, we pre-specified a set of inclusion criteria (Snyder, 2019), with the aim of defining the scope of our review. As implied by its name, the Uppsala model was developed at Uppsala University by Johanson and Vahlne, and has since been improved. Thus, our literature review focuses on the Uppsala model as it was developed by its authors over the years. We decided not to include papers from the authors that do not concern the Uppsala model or the development of its core concepts. Then, since this paper aims to rediscover the microfoundations of the Uppsala model, we maintained a strong focus on individual networks.

Moving on from these considerations, we conducted an advanced search on Business Source Complete/EBSCO combining the term “Uppsala model” with a number of key terms: “individuals”, “individual networks”, “individual level”, “micro level”, “micro-level”, “milli-micro level”, “nano level”, “nano-level”, “micro foundations”, “microfoundations”, “micro-foundations”, “nano foundations”, “nanofoundations” and “nano-foundations”. Indeed, there are several terms indicating the individual level of analysis, which required us to perform multiple searches. We then checked the correspondence to the authors. To better define the focus of our analysis, we first looked at articles in academic journals included in the most recent FNEGE ranking. However, we soon realized that the number of articles obtainable through this technique was very low (only five results), and that they were all quite recent. On reading the articles, however, it became clear that the relation between individuals and networks emerged in many papers that were not included in our EBSCO review.

This discrepancy can be explained by several factors. First, EBSCO, like other aggregators, checks the text of the article to search for specific words, but if the full text is not available, it simply checks the title, keywords and abstract. Some articles advancing the knowledge on the microfoundations of the Uppsala model were excluded because none of the terms relating to individuals were in the title or among the keywords (e.g., Wu & Vahlne, 2022).

Second, several individual features (including networks) were introduced into the Uppsala model from the social sciences. These features were not seen as an outcome of the Uppsala model, but rather as an explanatory tool to introduce the concept in IB. The lack of results could therefore be partially explained by a sort of natural outcome bias (Kahneman, 2011) of the EBSCO review, which highlighted the pattern “Uppsala model 🡪 individual” but not the reverse path “individual 🡪 Uppsala model”.

Third, the Uppsala model was created more than 45 years ago. Since then, the classification of reviews has emerged, rankings have been developed, and some journals have ceased publication (e.g., Journal of Market-Focused Management). Moreover, research books and chapters were more relevant than today, and novel topics were usually published in minor journals (Renwick et al., 2019). Our first review was thus distorted by a sort of ex-post cognitive bias (Kahneman, 2011): many results were excluded because they were either published in journals in other fields, or in journals that no longer exist or whose ranking has changed, or in research books.

In order to solve these problems, we further expanded our literature review (Booth et al., 2016). To make it systematic, we combined an author-based search on EBSCO with a cross-reference check of authors’ publications in their CVs and on Research Gate. We careful studied each article, summarizing all the sources in a table, including the main features of the article. We then selected publications according to the previously mentioned criteria. To avoid ex-post cognitive bias, we included academic sources such as books, chapters and journals that are out of print or no longer exist.

Ultimately, we selected 54 academic sources including 46 articles, two research books and six book chapters published over a period of 48 years, from 1975 to 2023. Following the recommendations of Tranfield et al. (2003), we classified the articles, extracting the main topics and elaborating a further synthesis. The classification is presented in the annexes and provides a summary of the academic sources for our literature review by year of publication (Annex 1). It includes the authors, year, title, journal or publisher, the main focus of the paper and a summary of its results.

Our review focuses on a specific topic within an evolving model, and we observed that the link between the model itself and the role of individuals has also changed over the years. To capture this evolution, we structured our review following a linear, longitudinal approach from the origins of the Uppsala model to the present days (Figure 2). This methodology is indicated to present the results of a systematic review and is extensively used in management sciences (Tranfield et al., 2003), as it enables researchers to explain the reality, interpreting events from a longitudinal perspective, avoiding the ex-post confirmation bias (Kahneman, 2011).

According to Vahlne (2021), even though the model was built through an incremental process, its evolution can be divided in three phases: origins, internationalization and globalization in networks, and evolution. In our literature review we stick with this structure but add a fourth phase focusing on the role of individuals.

Rediscovering the role of individuals within the Uppsala model

Origins and foundations of the Uppsala model

The Uppsala model was first developed to study internationalization processes. Two milestone articles are at the origin of the model. The first proposes a step-by-step model, known as the “establishment chain” (Johanson & Wiedersheim-Paul, 1975), where the internationalization process goes from the development of non-regular export activities to the establishment of foreign subsidiaries. The second, the actual Uppsala model (Johanson & Vahlne, 1977), focuses on knowledge development and commitment, the true engine of the establishment chain (Figure 3).

Figure 3

**The Uppsala internationalization process model**

Since its origin, the model developed around three main features: a process ontology, a behaviorist approach, and management under uncertainty (Vahlne, 2021). The first feature concerns the goal of the model: explaining how processes evolve over time. The two others concern the way companies and international development are conceived. From the outset, it was clear that this model differed from those based on the dominant neoclassical paradigm. As the authors themselves recall, “it seemed that Swedish companies did not behave as expected judging from textbooks in international business and economics. […] none of his four case companies had collected any data on the market before setting up sales subsidiaries. […] Rather, they tried one alternative mode after the other as knowledge was gained through experience” (Vahlne & Johanson, 2014, p. 160).

Far from being rational entities provided with complete information, companies focus on reducing risks. The main obstacle they face is psychic distance (Johanson & Vahlne, 1977), resulting in a risky environment where decisions are taken under conditions of uncertainty. Companies overcome psychic distance by increasing their level of commitment and market knowledge (Johanson & Vahlne, 1977) and begin their international development targeting markets with a geographical or cultural proximity with the countries of origin.

The establishment chain and the original Uppsala model were designed without any reference to individuals. However, the model is grounded on behaviorism, a theory of the firm that enabled it to accommodate the concepts of network and individuals (Vahlne, 2021). First developed by the Carnagie School (Cyert & March, 1963), behaviorism supposes that a company behaves, through aggregation mechanisms, as “a coalition of managers, workers, stockholders, customers and others, each with their own goals. […] Within the coalition some members exert greater influence and make greater demands for policy commitments than others. Such commitments, once made, become stabilized in the forms of budget allocations” (Miner, 2006, p. 61).

This approach reflects a micro-foundational perspective where decision making plays a key role, as in the case of individuals “Decision making is the heart of administration and […] theory must be derived from the logic of psychology and human choice” (Simon, 1947, p. XI). Thus, even if the Uppsala School focused on the organizational level, from the outset they developed “a model of knowledge development and increasing foreign market commitment” (Johanson & Vahlne, 1977, p. 23) in which companies learn, decide and face risks, in the same way as individuals do.

Integrating networks within the model

Although the original Uppsala model was a great success, it soon attracted criticism. Some critics argued that the model viewed internationalization as a linear, deterministic process (e.g., Knight, 1997; Meier & Meschi, 2010; Oviatt & McDougall, 1997; Tapia Moore & Meschi, 2010); others emphasized the model’s lack of adaptation to service companies and the fact that it did not take networks into consideration (e.g., Coviello & Munro, 1997; Forsgren & Johanson, 1994). In response to these criticisms, the model evolved by developing or integrating new concepts.

The model was extensively transformed by the introduction of the network approach into its theoretical framework (Johanson & Mattsson, 1987). Initially developed in mid-1970s by the IMP (Industrial Marketing and Purchasing) group, an international research network leaded by the Uppsala scholar Håkan Håkansson, the network approach became very influential in Sweden (Håkansson, 1982). The authors of the model quickly joined the discussion around this concept and contributed to its development.

The network approach is based on various considerations. First, connections between companies are decisive for competitiveness (Johanson & Mattsson, 1987). These connections take the form of an exchange relationship between customers and suppliers where trust, mutual commitment and learning play a key role (Anderson et al., 1994; Blankenburg Holm et al., 1999). Moreover, the network approach considers markets as networks where each company holds a network position (Johanson & Mattsson, 1985).

The concept of network soon became a core element of the model. In an article from 1987, Sharma and Johanson stress the role of networks for successful internationalization of Swedish consulting companies (partially answering criticisms about the lack of adaptation to service companies). From that point onwards, apart from a few exceptions, the goal of the model became to explain internationalization in networks (Forsgren & Johanson, 1994).

In the context of our review, the introduction of the network approach is of primary importance. Indeed the concept of network was developed in the social sciences at the individual level (Cook & Emerson, 1978; Granovetter, 1985), but used at the organizational level as a tool to explain customer-supplier relationship. Johanson & Mattsson (1987, p. 40) explain: “We have discussed interfirm relationships without explicitly referring to individual actors. However, the mutual orientation among firms is principally a mutual orientation among individual actors in those firms […]. Correspondingly, the interaction processes are carried out by individuals, though we have discussed them as taking place among firms”. The individual dimension entered the model not as an area of study, but as a tool to transfer the concept of networks into the model through the aggregation principle (March, 1962).

Drawing on the business network view (Forsgren, 2008), the model developed the concept of MBE – Multinational Business Enterprise (Vahlne & Johanson, 2013) – conceived as a network within networks (Håkansson & Johanson, 1993). This theory reinforced the anchoring of the Uppsala model in both the behaviorist theory and the network approach, both of which were originally conceived to study individuals and enabled the authors to develop in various directions. One research stream focuses on dyadic business relationships and adaptation in networks, introducing new concepts such as network position and network power independently of the Uppsala model (Anderson et al., 1994). A second steam uses the concept of network to reshape the model and respond to some of its criticisms. This stream divides the conceptual framework into two parts: on one hand, the establishment chain, on the other, the Uppsala model as an engine explaining the mechanisms of internationalization (Forsgren & Johanson, 1994; Johanson & Vahlne, 2003). In this context, the Uppsala school partially abandoned the establishment chain, refocusing on networks and mechanisms of internationalization such as knowledge development, learning and commitment (Forsgren, 2002; Johanson & Vahlne, 2003, 2006). In this way, they partially answered the critics regarding determinism and the lack of adaptation for explaining the behavior of born global companies.

The model was extensively transformed by the concept of business network, but also by the concept of opportunity, conceived as the result of exploration and exploitation activities and closely linked to knowledge development processes (Hohenthal et al., 2003). Developed separately, this concept was then integrated into the Uppsala model. In 2009, a revised model incorporating all these changes was introduced (Figure 4).

The new model considers that a company entering foreign markets is mainly threatened by the liability of outsidership (Johanson & Vahlne, 2009) – the disadvantage experienced by outsiders facing insiders in a specific network – which is a cornerstone in the evolution of the Uppsala school. By focusing on international network development, the authors were able to describe the behavior of service companies (Sharma & Johanson, 1987), and analyze uncertainty and commitment (Figueira-de-Lemos et al., 2011).

Alongside this, they developed a research stream investigating globalization in networks (Vahlne et al., 2011). This stream describes the tortuous road followed by companies in becoming global, introduces the concept of network reconfiguration and coordination for addressing issues such as headquarters-subsidiary relationships (Vahlne et al., 2012), and develops a “performance” state aspect in the Uppsala model.

To sum up, the concept of network was introduced into the model from the social sciences (individual level) and was turned into an engine to explain internationalization at the organizational level. Although the individual level was occasionally mentioned, the black box remained closed. However, the authors always recognized the existence of this shortcoming and held that “the company is clearly seen as consisting of a number of sub-units and individuals” (Johanson & Vahlne, 2006, p. 171).

Evolution in networks and evolution of the model

In 2013 the Uppsala model shifted its focus from internationalization to evolution in networks for the first time (Vahlne & Johanson, 2013). Since then, this framework has incorporated and enhanced multiple pre-existing features to expand its range of application. For instance, the authors further investigated the relationship between network knowledge and business relationship value (Hohenthal et al., 2014); introduced ambidexterity into the model as a dynamic capability (Teece, 2007; Vahlne & Jonsson, 2017); and extended the scope of the model to explain entrepreneurial processes through the business network view (Forsgren, 2016). Such concepts, deeply embedded in the individual level of analysis, contributed to the emergence of the role of individuals in the model.

The new model (Figure 5) aims to describe evolution processes “with internationalization and globalization being particular examples” (Vahlne & Johanson, 2020, p. 6). The term “network” no longer appears because it describes the nature of the MBE. Commitment and knowledge development processes, as well as capabilities (both operational and dynamic) and performance, take place within networks. The model looks back to its origins in conceiving company development as a process, but it is not deterministic: it does not make any reference to the stages to be taken or their speed of deployment but analyzes evolutionary mechanisms.

Figure 5

**The Uppsala evolution model**

During the same period, the authors started thinking about the evolution of their model over the years. In summarizing its main applications and clarifying its theoretical positioning (Holm et al., 2015; Vahlne & Johanson, 2014) they underlined the main challenges after 40 years. Among these, they identified the development of a multilevel approach as a major challenges for their model (Vahlne & Johanson, 2020). In this context, Vahlne & Johanson (2017, p. 1089) recall one of the original standpoints of their model, i.e., the aggregation principle, explaining that:

processes exist on multiple levels. The Uppsala model operates at the level of the individual firm, that is, the micro-level. When we record changes at the micro-level, they are to a large extent the aggregate outcomes of processes at the mille-micro level, i.e., the level of individuals or of subgroups within the organization. We have mostly treated the mille-micro level as a black box […]. We use the concept of evolution as the sum of changes at the mille-micro level, but aggregated to the level of the firm, where applicable.

As Vahlne (2020, p. 242) explains: “we were aware of the potential impact from micro-foundational factors, such as the individual managers, but stayed with studying the [organizational] factors”. The authors also underline the fact that they had included micro-level assumptions in the model, though without describing the role of individuals in the process (Vahlne & Johanson, 2017). At this point, the black box was ready to be opened.

Looking for the microfoundations of the Uppsala model

Actually, the black box was occasionally opened before 2017, but then carefully closed again. In 2010 Schweizer et al. (p. 365) proposed that “the revised Johanson and Vahlne model may be so general that it makes more sense to see it as an explanation of entrepreneurial change. In that case, the change may or may not imply internationalization”. Such thinking anticipated not only the focus on evolution, but also a new, multilevel approach. Aiming to explain internationalization as an entrepreneurial process, the focus of the paper was, for the first time, on managers (i.e., individuals) deploying dynamic capabilities to identify and develop opportunities. However, Schweizer et al. (2010) did not ultimately develop a multilevel model, but instead used individual features to enrich the Uppsala model. Choi and Johanson (2012) also focused on individuals by exploring the role of expatriates in knowledge transfer between HQs and subsidiaries. The authors concluded that, by developing personal relationships, expatriates enhance knowledge transfer – an important conclusion that was further developed later.

The explicit intention of developing a multilevel model emerged around 2017, following a general call for studies on microfoundations in IB (e.g., Contractor et al., 2019; Galkina & Chetty, 2015; Teece, 2007). The microfoundations theory argue that “to fully explicate organizational anything – whether identity, learning, knowledge or capabilities – one must fundamentally begin with and understand the individuals that compose the whole, specifically their underlying nature, choices, abilities, propensities, heterogeneity, purposes, expectations and motivations” (Felin & Foss, 2005, p. 441).

In particular, several authors invited the Uppsala model to adapt to the modern world by taking into account the role of macro-context and microfoundations (Figure 6), explaining that “ultimately, it is the individual who, through entrepreneurial action, connects various parts of the organization and the environment, and transforms opportunities into outcomes” (Coviello et al., 2017, p. 1156).

Figure 6

**The Uppsala model as a multi-level model**

To respond to the challenges raised by the academic community, the Uppsala school started working on a multilevel model. On the one hand, they tried to address macro-context issues (Bhatti et al., 2022) and, on the other, they focused on individuals (Vahlne & Schweizer, 2022). An initial article provided a model to describe managerial behaviors under uncertainty (Vahlne et al., 2017). In this paper, the Uppsala model no longer relies on micro-level concepts to explain meso-level phenomena but does exactly the contrary: it focuses on managers to understand how they influence the organization’s commitment decisions, learning, creativity, trust-building, and opportunity development processes. Several concepts used to define the characteristics of managers were already embedded in the model: management under uncertainty, bounded rationality, experiential learning, effectuation processes, opportunity development, network development. Others, such as emotion, vision and risk-taking were new.

The authors reaffirm their focus on managers in introducing the micro-level of analysis into their paradigm: “we foresaw radical change in the original version of our model […] relating modern psychological findings to the assumptions of the Uppsala model” (Vahlne & Johanson, 2020, p. 8). That same year, they explained how biases and emotions affect decision processes (Vahlne, 2020). The concepts of governance and coordination themselves were reinterpreted as network mechanisms that enable managers to build trust and, through other knowledge development processes as well, reduce uncertainty (Vahlne & Johanson, 2021).

The authors recently summarized their findings on the role of managers. They first show that managers’ capabilities play a key role in a globally shifting environment (Bhatti et al., 2022), then summarize individual characteristics and further explore emotions and biases (Schweizer & Vahlne, 2022). Uncertainty and risk are reduced by managers through knowledge development processes and commitment is adjusted over time. The change variables of the model are reinterpreted through this lens. The authors also underline the fact that thanks to its focus on evolution and individuals, the model can explain rapid internationalization in a non-deterministic way.

The Uppsala model tries to develop a multilevel perspective gradually, starting from change aspects. Knowledge development processes are analyzed from both an individual and an organizational perspective (Jonsson & Vahlne, 2021; Vahlne & Johanson, 2021). Commitment processes, then, are considered as dependent on managerial decisions (Schweizer & Vahlne, 2022). Thus, dynamic capabilities are considered to be an outcome of change aspects processed at a micro-foundational level (Wu & Vahlne, 2022). In our opinion, the change aspects successfully internalized a multilevel approach. The same, however, cannot be said about state aspects. Nevertheless, the authors intend to integrate individuals into every aspect of their model, as Figure 7 shows.

Figure 7

**The Uppsala model for the meso- and micro-levels of analysis**

The approach proposed by Vahlne and Schweizer (2022) aims to integrate the micro-level into the model (Vahlne & Johanson, 2017) without changing its structure. Every dimension needs to be analyzed from both an organizational and an individual standpoint. With this project in mind, the authors leave two missions for IB researchers to accomplish: first, to include individuals in state aspects (Vahlne & Schweizer, 2022); second, to take into account different organizational levels such as “the corporation, sub-units, and individual managers and co-workers” (Wu & Vahlne, 2022, p. 670).

Back to the future: taking innovative research paths

Our review shows that a new orientation of the Uppsala model toward individuals is emerging but is still incomplete. In the following paragraphs, we underline the main contribution of our review and the key points to be included in further research (Table 1).

Table 1

**The main contributions of our review and some recommendations for future research**

The first contribution of our review is to go back to origins to determine whether (and how) it is possible to introduce individuals in the Uppsala model. The answer is a resounding yes. Rooted in behaviorism (Cyert & March, 1963), the Uppsala model depicts companies and individuals as actors with bounded rationality, taking decisions under conditions of uncertainty (Vahlne & Johanson, 2014). Individuals aggregate into organizations and play a foundational role in their evolution (Contractor et al., 2019; March, 1962). We believe that, to further develop the Uppsala model, scholars should always adhere to its theoretical standpoints (a process ontology, behaviorism and management under uncertainty), especially when addressing multilevel topics. This is especially true with regard to the macro-context (Coviello et al., 2017), where the dominant economic theories rely on a very different view of the individual and reject the concept of evolution (Veblen, 1898).

Figure 8

**Combining the Uppsala model with a standalone, multilevel model**

A second contribution concerns the degree of integration of individuals into the model. We propose two alternative paths, in line with the suggestions by Coviello et al. (2017). The first involves integrating individuals into each dimension of the model, as shown in Figure 8. The focus so far, however, has been limited to the change aspects of the model. The authors explain that they have set aside state variables “hoping they will be the object of future study” (Vahlne & Schweizer, 2022, p. 1553). Some efforts have been made in this direction (e.g., Valdemarin, 2021), but there is still much to be done.

An alternative approach involves developing a standalone, dynamic model linking multiple levels of analysis that can be combined with the Uppsala model (Figure 8), as happened with the concept of network that was developed separately and then integrated in the Uppsala model. Similarly, it might be interesting to develop a framework focusing on how the micro-level influences the meso-level and vice versa and to integrate this into the Uppsala model. This new construct should be dynamic in order to capture the continuous dialogue between the individual and organizational level and, to be compatible with the Uppsala model, should feature a process ontology and a behaviorist approach. Such a construct would improve not only the Uppsala model but also the microfoundations theory.

A third contribution of this review concerns the identity of individuals. The model currently focuses on a specific group of individuals within companies, namely managers (Schweizer & Vahlne, 2022; Vahlne & Johanson, 2021; Vahlne & Schweizer, 2022). The authors consider that “Any action or decision undertaken at the firm level develops from individuals that make up the firm (Coviello, 2015). Consequently, understanding the firm’s internationalization behaviour stipulates understanding that it is the entrepreneurial individual driving it” (Bhatti et al., 2022, p. 2). This restriction, however, reduces corporate action to the decision of a single or few individuals. A company is an association of individuals (March, 1962), including employees, team leaders, shareholders and stakeholders in general. Without the application of the aggregation principle, any micro-foundational approach will be fragile. Thus, we believe that further studies should pay attention to different types of individuals and their roles.

Likewise, our fourth contribution concerns the external network. With few exceptions, the model addresses only customers (mainly conceived as companies). This is quite logic, since the network approach was developed to study customer-supplier relationships (Anderson et al., 1994; Blankenburg Holm et al., 1999). However, a multitude of actors, such as competitors, agents, former employees, and shareholders interact with the MBE and influence its evolution in various ways. What happens, for instance, if a former employee joins a competitor? A complete micro-foundational approach should address the external network, not only the internal network. To this end, future studies could adopt a qualitative and longitudinal approach to focus on external actors.

Similarly, the aggregation principle enables the identification of intermediate units of analysis. At present, the Uppsala model focuses with few exceptions on individuals and companies. Over time – and this is our fifth contribution – it should also include intermediate entities. In the internal network, such entities take the form of teams, business units, departments, etc. In the external network, they may include alliances, trusts, etc. In this way a flourishing field is opened up to IB researchers, with future studies based on a truly layered ontology, shedding light on a variety of topics from the perspective of sub-units and their members.

Our sixth contribution consists of discussing the characteristics of individuals. Drawing on the social sciences, the Uppsala model focuses on managers’ characteristics including bounded rationality, emotions and biases (Kahneman, 2011; Vahlne et al., 2017). Their decisions are taken under uncertainty and depend on their capabilities (Schweizer et al., 2010). However, it is important to detail the characteristics of individuals in the Uppsala model, in the same way that the characteristics of MBEs were extensively portrayed (Vahlne & Johanson, 2014). In this context, recent IB papers suggest that it might be interesting to integrate bounded reliability into the model (Foss & Weber, 2016; Kano & Verbeke, 2015). Deeply embedded in behaviorist theory, this concept could have a major impact on the model, by redefining the concepts of commitment and trust-building. In the same way, “other nano-level phenomena beyond the characteristics of key individuals of the firm, such as power distribution among different stakeholder groups, should be included in [multilevel] studies” (Vahlne & Schweizer, 2022, p. 7). Another interesting approach would be to review how individuals are conceived in IB, entrepreneurship and other fields of management and the social sciences. This will result in a better understanding of the behavior of individuals.

Moreover, the way in which individuals develop networks should also be examined in detail. Although the authors implicitly refer to some key concepts from the social sciences, it might be worthwhile developing a multilevel network approach including concepts such as structural holes and network entrepreneurs (Burt, 2009) or strong and weak ties (Granovetter, 1973). How do individuals develop networks? Are they affected by belonging to the company’s network? How does the behavior of individuals in networks influence companies and organizations? Further studies could focus on multilevel network development processes from a multidisciplinary perspective.

We consider that the rediscovery of the role of individuals would have a strong impact on the Uppsala model. First, it would provide a definitive answer to certain criticisms, shifting the focus from progressive vs rapid internationalization to the mechanism of internationalization (Schweizer & Vahlne, 2022). Second, it would reinforce the orientation of the model toward evolution, of which internationalization is a sub-category. In addition, a stronger focus on individuals would result in the development of additional “managerial” outputs, making the model capable of spanning topics in IB through to international management. In this regard, the scope of the model would expand.

Alongside its theoretical results, our paper offers some methodological recommendations for developing systematic literature reviews. First, it is worth noting that aggregators (such as EBSCO) simply check the article’s title, keywords and abstract if the full text is not available. Thus, some articles may be excluded from the results. Second, it is important to overcome the outcome bias when conducting a systematic review. In our case, individual networks were not initially investigated by the Uppsala model but were used to introduce the concept of network in management sciences. Third, several factors such as the evolution of academic rankings or the closing down of a journal may reduce the number of results. Taken together, these various points led us to propose a set of methodological recommendations. To avoid biases (Kahneman, 2011) it is essential to consider multidirectional paths (e.g., “Uppsala model  individuals” rather than “Uppsala model  individuals”) and to pay attention to the previously mentioned contextual elements when developing a systematic review.

Conclusions

Our paper has aimed to rediscover the roles of individuals within the Uppsala model by developing a systematic literature review from its origin to the present day. Our results show that from the outset the model was based on multilevel standpoints. With the development of the revised model (Johanson & Vahlne, 2009) the authors integrated the network dimension from the social sciences, thus reinforcing the potential for a multilevel approach. In shifting the focus to evolution, the concept of network became a prominent part of the model. Finally, in recent years, the authors have tried to move toward a multilevel approach (Vahlne & Schweizer, 2022). Our analysis shows that, although several steps have already been taken in this direction, there is still much to do.

Our first contribution shows that the theoretical standpoints of the Uppsala model enable it to develop on multiple levels. However, recent papers have only analyzed the change aspects (Schweizer & Vahlne, 2022; Vahlne & Bhatti, 2018; Vahlne & Johanson, 2021; Wu & Vahlne, 2022) and have not addressed state aspects. Our second contribution is to propose two possible solutions to this omission. The first is to conceptualize the role of individuals in every aspect of the model, as proposed by Vahlne and Schweizer (2022); the second is to develop a standalone framework linking multiple levels of analysis that can be combined with the Uppsala model.

Our third contribution is to reaffirm the importance of the aggregation principle for developing a micro-foundational approach. By focusing solely on managers, there is a risk of oversimplifying the model. A similar limitation can be found with regard to the external network, where a micro-foundational approach has been little, if at all, developed. Our fourth contribution lies in providing indications to include individuals in the external network within the Uppsala framework.

Furthermore, the model also ignores intermediate units (departments, teams, etc.) within the company, thus oversimplifying reality. We propose – as our fifth contribution – that new research paths should include those actors in the model, thereby enabling the development of a holistic, comprehensive micro-foundational approach.

Our sixth contribution concerns the characteristics of individuals. The Uppsala model has recently included several characteristics such as emotions, biases and bounded rationality. However, other characteristics such as bounded reliability could be taken into account (Foss & Weber, 2016; Kano & Verbeke, 2015). Likewise, it would be interesting to focus on how individuals develop networks.

Our paper also presents a set or methodological contributions that may help researchers to conduct systematic reviews. We show how aggregators (e.g., EBSCO) check available material only. We point out the risk of incurring in an outcome bias (Kahneman, 2011) when researching new topics. Finally, we offer some considerations concerning the evolution of rankings and the creation and disappearance of journals and role of other academic sources that might be helpful for scholars delving back in time.

Together with contributions, our review has certain limitations. The first is our author-based approach. Although we have investigated other authors’ contributions, our review limits itself to the model developed by Johanson and Vahlne over the years. On the one hand, this model has been adapted by other scholars and the original authors have internalized the changes suggested (e.g., Coviello et al., 2017; Forsgren, 2016). On the other hand, some concepts that were developed outside the careful control of the authors could reshape the future of the model itself. This possibility leaves unexplored a potentially interesting topic: the evolution of the Uppsala model outside the Uppsala school.

Concerning levels of analysis, we have limited our investigation to the micro-level, setting aside the macro-context (Coviello et al., 2017). We are aware that the Uppsala model is currently trying to address this criticism and become a fully integrated multilevel model (Bhatti et al., 2022). Nevertheless, we believe that integrating the macro context would require an in-depth analysis of the theoretical standpoints of the model to determine with which economic theories it is compatible. A cross-disciplinary approach is recommended to investigate this topic. After 48 years, the Uppsala model is still ready to confront new IB challenges.