Bridging the gap between high-speed rail transport studies and cluster economics through social knowledge exchange: future research potential

ABSTRACT

Within the high-speed rail debate (HSR) there has been lacking in-depth theoretical and evidential research on the role HSR has on tacit knowledge flows between industrial clusters or regions of economic productivity. The research that has begun to emerge has focused on knowledge indicators such as patents for social/tacit knowledge exchange, which this paper raises concerns over. This review aims to discover new links for HSR with cluster economics through social knowledge theory and aims to create a conceptual framework that will provide a new perspective for future research in the HSR debate concerning social knowledge exchange. Potential implications are presented for future transport policy decision-making, based on the relationship HSR may have with inter-regional tacit knowledge flows and accessibility benefits in regional balancing. Future research considerations are pointed out that argue to measure the flow of knowledge from HSR, research needs to go in-depth to the social aspect of interactions and relationship buildings, as quantitative data will struggle to capture tacit knowledge flow itself, due to the tacit nature of such information. The presented framework highlights the tacit nature of interactions facilitated by HSR connectivity.

KEYWORDS:

• High-speed rail
• knowledge flow
• cluster economies
• agglomeration
• social knowledge theory

1. Introduction

In discussions about the costs and benefits of high-speed rail (HSR), concerns arise regarding the expensive development and operations of HSR infrastructure (trains surpassing 250kph). If it is determined that HSR promotes knowledge flow and innovation, and if HSR contributes to balancing regional development and economic growth, then these costs may be deemed justified. The cost–benefit debate has moved on towards exploring wider economic benefits (inter-city agglomeration economies), with a new wave of research investigating the impacts of HSR on knowledge flow, dominantly done by quantitative positivist approaches. This paper opens the debate on the implementation of HSR by aiming to provide a new perspective for future policy implications based on HSR and its involvement in social knowledge flows between regions housing industrial clusters.

Existing research often overlooks the nuanced dynamics between cluster formation, transportation accessibility, and the social flow of tacit knowledge. This paper offers a novel perspective for future research on the potential of HSR development in reshaping and fostering collaborative networks across regional clusters. This is carried out by illustrating the limitations of current studies in capturing the social dimension of knowledge transfer and emphasising the need for in-depth qualitative analysis. A conceptual framework is presented that indicates HSR as a catalyst towards inter-regional knowledge exchange, and thus regional innovative potential and development balancing.

HSR research and literature have grown significantly over the last couple of decades, in line with its development as a new transport backbone for many counties. Studies have explored diverse impacts of HSR, including time savings (e.g. Cascetta et al., 2020; Kim et al., 2019; Wang et al., 2019), to time–space convergence (Jiao et al., 2014; Spiekermann & Wegener, 1994) and population flow/potential (Hou & Li, 2011; Martín et al., 2004; Zhang et al., 2016), which collectively shape socio-economic landscapes across regions (Cheng & Chen, 2022). HSR catalyses economic growth (Chen et al., 2016; Li et al., 2020; Vickerman, 2018), through tourism (Bazin et al., 2011; Chen & Haynes, 2015, 2012; Wang et al., 2012; Zhou et al., 2016), employment (Chen et al., 2016; Dong, 2018; Hiramatsu, 2018; Venables, 2007; etc), output growth (Donaldson, 2018; Li et al., 2018; Liu & Zhang, 2018; Zhao et al., 2017; etc), and business journeys (Baum-Snow & Kahn, 2000; Ollivier et al., 2014; Zhou et al., 2016). HSR can bring regional equity in economic performance (Chen & Haynes, 2017; Hiramatsu, 2018; Jacobs-Crisioni et al., 2016) or the opposite and exaggerate disparities (Jin et al., 2020; Martín et al., 2004; Monzón et al., 2010). What is lacking in the breadth of HSR literature is exploring HSR impacts on tacit knowledge flows and innovation, through more tacit-nature conscious methodologies. This is particular for geographic agglomerations of industries that offer a wealth of knowledge that appears trapped in a bubble or struggles to be shared outwards due to its tacit nature and reliance on closed network ties.

Current literature does not adequately address the interaction between HSR and agglomerations, within the realm of cluster economics at the tacit dimension. Specifically, the transmission of tacit knowledge through face-to-face interactions via tacit focussed methodologies. Literature has begun to highlight the concept of HSR and “Knowledge Innovation”, with recent papers on HSR and knowledge spillovers (Hou, 2022; Inoue et al., 2017; Lu et al., 2022), exchange (Long & Yi, 2024; Wang et al., 2022), collaboration (Wang et al., 2022) and productivity (Bhatt & Kato, 2021; Komikado et al., 2021; Miwa et al., 2022) as well as research productivity (Dong et al., 2020). However, certain aspects, such as socially experienced interactions and tacit-based learning, are lacking which leads to existing domination in the research methodologies of knowledge flows undertaken in HSR research. This opens up an opportunity for new approaches catering for a deeper understanding of knowledge flows via HSR accessibilities.

HSR offers a novel new potential for not just individuals, but also the knowledge they carry, across inter-regional spaces that were previously deemed inaccessible for one-day business trips. HSR and industrial clusters are considered separate entities, and few studies strongly connect the two. Integrating the two bodies of literature is needed to understand the potential bridge they may have to each other. This review addresses this gap by constructing a literature review and conceptual framework to identify the relationship and benefits of future research. Section 2 aims to clearly explain the relevance of social knowledge theory and cluster economics towards transport and HSR studies, highlighting a potential partnership. Section 3 delves into the existing literature on HSR and knowledge flows, offering a critical assessment of current quantitative methodologies and approaches, along with an overview. This paper then provides a conceptual framework in Section 4, where a theory is proposed on how HSR enables multiple clusters to develop their knowledge economies by promoting the movement of tacit knowledge carriers. This is followed by concluding remarks highlighting future research needs. Therefore, this literature review aims to bridge HSR with cluster economics and create a conceptual framework offering a fresh perspective on the HSR debate.

2. HSR dynamics in social knowledge theory and cluster economics

2.1. Social knowledge theory

Understanding social knowledge theory is important as it offers a socially grounded perspective to explain cluster economics, elucidating the interactions and learning processes within such industrial hubs. Such interactions are what HSR would be able to exploit and expand. Chatterjee (2017) defined social knowledge as knowledge embedded in societies, with individuals learning within their communities through participation (Arrow, 1992). Organisational learning encompasses both individual knowledge acquisition and organisational participation through social processes (Sfard, 1998). Elkjaer (2004) highlighted a third way expressed by Strauss (1978) as the social world theory, where combined acquisitions by individuals and participation within organisations contributed to an experience of interactions. This was similarly put by Clarke (1991) as “Groups with shared commitments to certain activities, sharing resources of many kinds to achieve their goals, and building shared ideologies (p. 131)”. Elkjaer (2021) used this and Örtenblad’s (2018) versions of the learning organisation to describe the impact of experience as an “interactive way of living – and work” to which knowledge “may be used to describe, understand and elaborate”. Regarding the social world theory Strauss (1978) discussed, Shibutani (1955) had argued previously that such interchanges were bounded by the limits of effective communication. Therefore, with the rise of HSR accessibility, this review proposes a discovery of what is deemed modern limits/boundaries of face-to-face communications.

Expanding upon learning being a social process, this paper argues that HSR may redefine the boundaries of Wenger's theorised communities of practice. Here, “members of a community are informally bound by what they do together—from engaging in lunchtime discussions to solving difficult problems” (Wenger, 1998). Knowledge creation and sharing were based on the communities (the knowledge carriers and talent within a boundary). Through “mutual engagement”, a community of practice is “an intrinsic condition for the existence of knowledge” (Lave & Wenger, 1991; Wenger, 2005). Knowledge is gained through participation in “social learning systems”, where learning is an “interplay between social competence and personal experience” (Wenger, 2010). However, these knowledge theories have a common trait of focusing on singular communities or organisations and have not looked beyond local geographies. Tacit knowledge exchange through face-to-face interactions can now benefit from the accessibility benefits of time–space convergences (Jiao et al., 2014; Spiekermann & Wegener, 1994). With modern technology and transport, communities and organisations may not be considered as isolated knowledge bubbles, but also a part of larger more accessible one-day catchment areas (which make business interactions more compelling, efficient and less time costly). For example, HSR-induced one-day catchment areas (Kim et al., 2019) may warp boundaries out to other bubbles, forming new inter-regional metropolises of knowledge exchange and practice.

2.2. Cluster economics

With the importance of communities in the social learning ecosystem, Bathelt et al. (2004) discussed how clusters become catalysts for such communities, emphasising institution-building and day-to-day interactions within a cluster. An “Industrial Cluster” is defined as a localised agglomeration of economic and innovative activities where an inter-firm face-to-face network of business relationships exists. A cluster houses communities of practice, and the tacit knowledge and skills of individuals. The development of cluster infrastructure through intra-regional transport, accessibility, social infrastructure (networking events, conventions, etc.), industrial zones, public institutions, etc. (see Porter (1998)), allows the accumulation of knowledge in the sense of information, suppliers, a workforce and specialised R&D (Porter, 2001). Cluster economics dates back to Marshall (1890) who first described an air of knowledge existing within a cluster, due to the agglomeration of firms, a specialised labour force, and suppliers. Porter (2001, p. 19) furthered the concept by defining a cluster as “geographically proximate groups of interconnected companies and associated institutions in a particular field, linked by commonalities and complementarities”. Here, closely linked firms and individuals were able to communicate and share concepts easily without geographical or topical obstacles present. Such communications are key to the success of agglomerations (Malmberg & Maskell, 2002).

The face-to-face network of a cluster is a vital aspect of its success and its interactions offer great potential (Markusen, 1996). Bathelt and Taylor (2002) argued that clusters act as “temporary and unique stabilizations of relationships”, which push towards trust and closely-knit ties within competing yet cooperating firms. Trust increases interactions and relationship building within the cluster, reducing interaction costs and increasing agreements to share knowledge among firms, through “bonding social capital” (Malmberg & Maskell, 2002; Rutten & Irawati, 2013). This face-to-face knowledge bubbled within a cluster economy, creates what Boschma et al. (2017) called a local buzz, vital for a cluster to grow, compete, learn and innovate. By supporting the formation of a local buzz, face-to-face networks share social knowledge, supporting mutual innovation and pushing towards novel knowledge creation (Bathelt et al., 2004). Local linkages within a cluster aid competitive advantages from a system of shared values (Bathelt, 2005). The successful industrial cluster exists therefore as a powerhouse of productivity due to specific knowledge, suppliers, employees, R&D and competition within one localised closed area (Porter, 1998).

Learning in large and dense areas is highly efficient, as it facilitates frequent and diverse interactions among different groups, enabling individuals to connect with skilled neighbours and enhance their learning opportunities (Glaeser, 1999). This concept holds within cluster agglomerations, where knowledge creation thrives through dense and efficient interactions within a knowledge community (Henry & Pinch, 2000). Tacit knowledge is obtained through involvement, experience and witnessing of practices and occurrences that cannot be learned through codified means such as media, telecommunications and emails. It is an “interactive learning process” (Maskell & Malmberg, 1999) which is most efficiently picked up in localised areas, face-to-face. This is what Von Hippel (1994) called sticky: knowledge or information stuck to one place where it grows and is shared, while codified easily gets out and shared among the greater population. It is the sticky spatial proximity of tacit knowledge that makes it so valuable and hard to obtain elsewhere, which mixed with its social dynamism and community-based approaches, creates a powerful combination for high-speed rail to potentially exploit.

Systemic learning and interactive innovation strengthen the overall regional innovation system, strengthened by trust and cooperation (Cooke et al., 1997). Face-to-face interactions encourage communication, commitment, screening of actors as well as competition (Storper & Venables, 2004). Cooperation formed through these processes permits the sharing of tacit knowledge in the local knowledge community (Lawson & Lorenz, 1999). Innovations occur more frequently in well-linked agglomerations (Cook et al., 2007) and they allow for innovations to occur quickly at low costs (Porter, 1998). Feldman and Audretsch (1999) found that there was a relationship between innovation and firms within local proximity that were a part of the same common science base. Boschma et al. (2015) used relatedness to highlight similar, that new technologies were likely to emerge in a city when related to current technology, due to the existing knowledge bubble of the city. This knowledge bubble is defined in this paper as an invisible boundary of localised available tacit knowledge “in the air” (Marshall, 1890). In other words, in a cluster of firms that rely on similar R&D (knowledge development), although consisting of diversified industries, innovation occurred at a larger scale due to the close geography of organised knowledge. Policy and programmes focusing on interactions and communications allow for the knowledge community of a cluster to efficiently pursue innovation (Castells & Hall, 1994), yet there lacks much literature on how external infrastructure policies may influence such goals.

The free flow of knowledge between firms, and collision/sharing of ideas and methods, is described as open innovation, which encourages knowledge convergence (Yun et al., 2016). Hacklin et al. (2009) defined knowledge convergence as “the emergence of serendipitous coevolutionary spill-over between previously associated and distinct knowledge bubbles, giving rise to the erosion of established boundaries that isolate industry-specific knowledge (p. 725)”. Knowledge convergence is the merging of two different knowledge bubbles that allow for two previously indistinguishable industries to cooperate and innovate. This may occur through both formal interactions such as work meetings and partnerships, or informal such as dinner or drinks with colleagues from the area. Open innovation exploits external tacit knowledge accessibility (Park, 2017) and ensures an environment of creativity and innovation. This means open innovation is only to benefit from intermediaries that encourage knowledge networks and provide localities with new opportunities to share knowledge (Howells, 2006). The literature above has focused on singular localities of clusters that maintain a network within themselves. How external infrastructure, like HSR with time–space shrinkage, turns geographies that existed previously as two separate regions, into one larger catchment area should ring alarm bells as an exponential potential in innovation and knowledge growth. Creating an inter-regional metropolis in knowledge bubble form through high-speed rail connectivity is a concept to investigate further, to which this review aims to construct a framework for and critique.

3. Current work between high-speed rail and knowledge, a potential partnership?

3.1. HSR literature and knowledge flow

Throughout history, transport has been shown to enforce and encourage urbanisation and agglomeration of industries, leading to city growth (Chen, 2016; Clark, 1958; Teaford, 1986). Trade increases when transport costs are reduced and made more accessible, and when paired with improved services (such as speed and comfort), transport can encourage agglomeration benefits (Chatman & Noland, 2011; Graham, 2007a. Focusing on labour demand (productivity and wages), Venables (2017) measured the impact of inter-city transport improvement, finding reductions in communication costs between cities would improve labour supply and income, with reduced trade costs also increasing task specialisation. Moomaw (1983) analysed transport infrastructure to population demography and found that transport benefited productivity within agglomerations, while Graham (2007b) measured transport investment in agglomeration economies, finding a 10–20% increase in economic return. Clusters benefiting from transport investment, have increased effective density (spatial arrangement and accessibility of individuals and activities), output per worker and firm output (Graham, 2006; Jin et al., 2014). Current literature within the realms of transport studies support the argument that transport aids agglomeration economies through improved accessibility and improved time–space population densities at an inter-regional level. Most of these studies are also adopting quantitative ex-anti modelling approaches and it is not clear how such HSR impacts work for particular industries. How HSR serves a greater inter-regional landscape is worth expanding upon.

Knowledge production is an emerging field in HSR studies. Knowledge diffusion has been argued to be influenced by geographical distances (Boschma et al., 2015; Peri, 2002), influencing knowledge-intensive activities (Chen & Hall, 2011). Geographic distances are understood to be shrunk through the time–space convergence of HSR (Jiao et al., 2014; Spiekermann & Wegener, 1994). In Spain, Matas et al. (2020) found (using panel data) that HSR had a larger effect on “knowledge-intensive activities” and services, less so on manufacturing, proposing it may promote firm creation where face-to-face relationships are needed. Charnoz et al. (2018) researched the impacts of HSR in France, concerning communication costs, finding a decrease in communication costs as a result of HSR connectivity increased soft information transmission.

The transformative potential of transport projects in reshaping regional economies maintains importance when considering these dynamics in policy decisions and future research efforts. Graham (2006) emphasised the significance of agglomeration in influencing productivity, particularly in urban economies, and highlighted the crucial role of transport infrastructure when shaping such dynamics. Furthering this by suggesting wider economic benefits from transport investment through changes in densities (transport costs) available to firms (Graham, 2007a). Venables (2017) expanded on this by arguing that large transport projects, such as high-speed rail networks, can have substantial impacts on economic geography and regional activity structures. Venables demonstrated how both inter-city and intra-city projects can generate wider benefits by improving business links and facilitating access to goods and services across cities, promoting city specialisation and welfare gains. Chen and Vickerman (2017) added to this discussion by emphasising that new HSR projects impact regional economies by aggregating labour markets rather than solely extending single labour markets. Chen and Vickerman suggest that while there will be agglomeration effects, the transformative potential and regional rebalancing resulting from such projects may outweigh the direct agglomeration effects. With this it was put forward the question of how new transport investments can alter the economic situation of cities or regions, highlighting the need for further analysis in this regard. This paper therefore builds upon this narrative of exploring and expanding how inter-regional knowledge bubbles work by providing the next step in HSR literature and methodologies, beyond correlation and regression analyses, while appreciating current efforts in quantifying such impacts.

HSR investment is realised to a large extent in highly developed urban agglomerations, suggesting HSR is often a source of reinforcing development, not initial development, to areas that have reached a high level, or even plateau, in economic productivity (Chen et al., 2016). This can suggest that high-tech clusters may reach a point where growth and expansion are limited without the addition of forms of infrastructure that will aid in additional networking and accessibility of knowledge and firms. HSR influence on knowledge spillovers, labour market pools and efficiency of linkages, as argued by Graham and Melo (2011), can influence agglomeration economies, leading to growth in denser areas (Shao et al., 2017). Literature has begun to rise on the concept of HSR and “Knowledge Innovation”, with recent papers on HSR and knowledge spillovers (Hou, 2022; Inoue et al., 2017; Lu et al., 2022), exchange (Long & Yi, 2024; Wang et al., 2022), collaboration (Wang et al., 2022) and productivity (Bhatt & Kato, 2021; Komikado et al., 2021; Miwa et al., 2022) as well as research productivity (Dong et al., 2020). The accessibility and socio-economic benefits that HSR brings between regions could act as a new form of knowledge transfer medium.

Current work to investigate or argue towards HSR influencing knowledge production and sharing has followed multiple routes. First, Dong et al. (2020) researched the influence of HSR on knowledge creation across cities using academic paper co-authorship. This was through the encouragement and uptake of face-to-face interactions and improved teamwork of researchers from different cities. The argument put forth was that HSR accessibility improvements from increased travel speed mean face-to-face activities cost less between skilled workers. Results showed co-author productivity increased as well as new pairings. The most productive influence HSR had on researchers were those from social science backgrounds. Such correlations do not necessarily show true social knowledge flow and can represent other factors such as improved funding, development, or research networking events. Publications may act as an indicator but seldom demonstrate any official interactions based on HSR connectivity.

The second and most common method to investigate knowledge sharing through HSR systems has been co-patenting analysis. This methodology has argued an increase in co-patents post-HSR construction indicates more networking between inventors of two or more locations. Long and Yi (2024) showed that HSR networks increase patent citations, arguing a demonstration for increased knowledge exchange between regions, within a 4-hour travel time. Hou (2022) investigated HSR-induced urban access to different innovation factors through patent analysis, suggesting significant benefits to knowledge bubbles and industrialisation specialisation, especially for over 2-hour travel times. Miwa et al. (2022) used municipal-level panel data to argue that HSR provided positive effects on regional innovation due to implied inter-regional communication opportunities. Inoue et al. (2017) similarly found communication improved after the Nagano-Hokuriku Shinkansen line connected Tokyo and Nagano in 1997, fostering innovation and collaboration among firms. Wang and Cai (2020) identified that less-developed cities connected through HSR participated in significant innovation improvements, finding connections to highly-developed cities and aiding research collaboration. This was supported by Du et al. (2020) who found city access in China to HSR increased their innovative performance, most noticeably in knowledge-intensive industries. Yao and Li (2022) found that co-patenting, patent quality and collaborative partnerships increased with HSR connectivity, most significantly between city pairs under 250 km. Xiao et al. (2022) also used patents to find the nation’s HSR network increased intercity technology transfer, noting that HSR in China had increased the likelihood of face-to-face interactions as a result of increased accessibility and connectivity. These findings were supported by Wang et al. (2022) who identified that cross-city innovation collaboration significantly rose post-HSR and those cities connected with HSR showed noticeable positive growth in collaboration compared to those without. These findings were also correlated with travel time between city pairs and argued that investing in high-speed transport increased knowledge flow channels.

Table 1 provides a summary of the majority of the literature mentioned, and highlights the recurring themes to measure knowledge-sharing activities being based primarily on quantitative methodologies, particularly patent analysis, data sets, and panel data.

Table 1. Summary of current HSR literature focussing on knowledge-sharing activities.

	Countries	Focus	Data
Charnoz et al. (2018)	France	Communication cost	Geographical, Internal Workforce, Accounting Indicators
Graham and Melo (2011)	United Kingdom	Agglomeration benefits	Commuting Flow Data (National Transport Model), Proposed HSR Data (HS2), HSR Data, National Statistics (GDP)
Wang et al. (2022)	China	Labour Flow	Prefecture-level, Patents, China Migrants Dynamic Survey
Matas et al. (2020)	Spain	Knowledge-intensive activities	NUTS-3 Regional Data, Central Business Register Statistics
Lu et al. (2022)	China	Knowledge Spillovers	City Level Economic Indicators (e.g. Population and investment density, GDP, Entrepreneurial Activity)
Komikado et al. (2021)	Japan	Knowledge Productivity	Patent, Dichotomous Dummy Variables, Effective Density, Firm/Employment Data
Miwa et al. (2022)	Japan	Knowledge Productivity	Socioeconomic and Sociodemographic Indicators, Patent, and Transportation Variables
Hou (2022)	China	Knowledge Spillovers	Patents, R&D Inputs and Employment Statistics (Census Data)
Inoue et al. (2017)	Japan	Knowledge Spillovers	Patent Citation Relationship, Patent Submission Data
Long and Yi (2024)	China	Knowledge Exchange	Patent Citations, HSR Data, Regional Statistics (GDP, Population, Passenger Volume, Internet Users)
Wang et al. (2022)	China	Cross-city Innovation Collaboration	Co-Patents, HSR Data, City Socioeconomic Indicators (e.g. Population, Ridership, Internet)
Bhatt and Kato (2021)	Global (up to 59)	Knowledge Productivity	Panel Data (Patent, GDP, Urban Population Growth, Researchers, Graduates, etc)
Yao and Li (2022)	China	Collaboration	Co-Patents, Economic Indicators (e.g. GDP, Science and Technology Government Expenditure, etc)
Xiao et al. (2022)	China	Technology Transfer	Transaction Data on Intercity Patent Rights, HSR Data, Economic Indicators (E.g. Population, PGDP)
Du et al. (2020)	China	Innovative Performance	Patents, HSR Data, City Data (Population, GDP, etc)
Dong et al. (2020)	China	Research Productivity	Research Paper Publication and Citations (WoS), Economic Indicators (University Investment, Passenger Volume, GDP, Population)

Source: Summarised by authors.

Although this current literature points out the role of HSR in facilitating face-to-face interaction and enhancing knowledge economies, they do not offer further insight into how knowledge creation has worked at the local levels and how the role HSR has played in facilitating knowledge interactions. If such HSR impacts exist, it would be a significant contribution to the HSR cost–benefit debate. Knowledge goes far beyond codified means and there lacks literature on the social front of knowledge exchange. Using patents or co-authorship, as well as other empirically based metrics, undermines the social nature of knowledge exchange, and relying on such correlations does not offer depth in answers. Social learning systems innovate through “active boundary processes” (Wenger, 2010), engagements within the boundary of a community of practice, and HSR expanded one-day catchment areas and time–space warping can be examples of merging boundaries. Current work on knowledge exchange through HSR networking and development has tended towards correlation and codified data points with suggestive implications of increased interactions. Without digging into the social aspects of knowledge creation and flow and investigatory discussions, any face-to-face interactions or accessibility gains from HSR between two or more individuals cannot be confirmed, hence the importance of understanding social knowledge theories before these studies. Social knowledge exchange may result in patents or research papers; however, patents and papers are not only formed from social knowledge exchange. There is a research gap on how social knowledge theory is linked to the operation of inter-regional face-to-face accessibility from high-speed transport, both at the individual and community levels. This review contributes to the deeper understanding and future research of the relationship within inter-regional face-to-face induced HSR accessibility, as this can remould how inter-regional development and transport planning are perceived.

Cooley (1926) in “The Roots of Social Knowledge” stated that “Statistics is an exact method, and it is enabled to be such precisely because it is not in itself social but mathematical. It does not directly perceive social facts or any other kind of facts … (p. 73)”. We propose a research topic that follows this mode of thinking, regarding the current argument on HSR and knowledge exchange. We agree with Gurukkal (2019) that “the study of society makes sense only if we conceive it in terms of its constituents, relations, institutions, ideas, practices, structures, systems, and processes (p. 23)”. And as put by Stark (1950);

Social knowledge cannot be secured […] by applying the methods of the exact sciences, because all their knowledge is necessarily outside knowledge, knowledge “about” something, not “acquaintance with” something, as all true sociological perception is. We would, therefore, plead with philosophers to widen the basis of their epistemological speculation by including in it the hitherto neglected findings of the social sciences. (p. 307)

The rise in knowledge studies in HSR literature is positive, but this research review proposes investigating the truer relationships of what knowledge is, how it moves, and how HSR acts as a bridge to such desired exchanges if it indeed does. We express concerns over the majority of research entering this field, as the quantification and measurement of knowledge have focused on codified means and correlations that assumed tacit face-to-face exchange has taken place.

3.2. The potential partnership

HSR has the potential to bring newcomers at a greater scale into participation with the community of practices that host a mastery of knowledge in other bases (Lave & Wenger, 1991). During discussions and research in knowledge exchange by the sociologists of social knowledge theories and practices, singular localities were never discussed on an inter-regional scale, as transport technology was never particularly present to the accessibility levels required for inter-regional warped proximity. HSR’s potential to create larger knowledge bubbles raises questions about the definition of boundaries within communities of practice or clusters when inter-regional accessibility increases. Are HSR-connected clusters singular localities or multiple? And does HSR reshape spatial proximities within innovative industries, prompting talent/knowledge carriers to identify themselves within a region or as part of a larger inter-regional metropolis?

Tacit knowledge relies heavily on face-to-face interactions to allow its transfer to other professionals who adapt and utilise it to innovate (Polenske, 2007). Individuals need to be within close proximity to allow for face-to-face contact to occur, such as clusters. Physical interactions cannot be replaced due to the difficulty of knowledge conversion turning what is tacit into codified, as indescribable information is lost (Nonaka, 2000). This review assumes that tacit knowledge drives innovation within agglomerations, where the socialisation of professionals and knowledge contributes to cluster success and innovation. What lacks is a firm foundation and framework towards how transport can exploit and improve upon this closed system of clusters, to enable tacit knowledge to be brought out of the close spatial proximities they are stuck to, allowing face-to-face interactions to occur between individuals of different regional localities.

HSR provides the growth in transport accessibility that offers a new pathway for social tacit knowledge carriers to reach new clusters previously deemed inaccessible, particularly 200–800 km away or 1–4 h of travel (Givoni, 2006; Hall, 1999; SDG, 2004). When door-to-door travel time is below 2 h (or city-centre to city-centre), trains are considered as fast or faster than air travel, which HSR can offer within new geographical ranges compared to conventional rail, while HSR maintains a market split with air between 2 and 6 h (UIC, 2023a), HSR offers a uniqueness for inter-cluster connectivity over air and road transport beyond lower travel time, but also in comfort, price, reliability, accident reduction as well as extra capacity (De Rus, 2009; Givoni, 2006; UIC, 2023b). Regarding capacity, HSR can theoretically maintain the same number of passengers in a Boeing 737 every 45 s, or the equivalent of three parallel motorways and a 4 km long line of cars (SDG, 2004; UIC, 2023a).

HSR also offers reductions in congestion both within the capacity restraints in airports, as well as on roads, due to the increase in capacity and freeing of capacity on conventional transports (Givoni, 2006; SDG, 2004). Following the introduction of HSR, De Rus and Inglada (1997) found that in both air transport and road usage between Madrid and Sevilla, air demand dropped 50%. HSR offers a decrease in waiting time on both ends of air travel, demonstrating superior access and egress (Morrison & Winston, 2005), and environmental benefits (Givoni, 2006). An example of these benefits is demonstrated in Taiwan’s HSR, decreasing travel from the North to the South from 4 h to 1.5–2 h, and declining road/air traffic as a result of a new “one-day peripheral circle” (Kim et al., 2019; Li et al., 2015). An additional benefit to consider is the improved work productivity of laptop/mobile usage on rail in general, in comparison to air and road travel. HSR in these ways offers a unique expansion of previous boundaries in the one-day catchment model of previous business and public travel.

Clusters and HSR infrastructure share an ability to move knowledge within their respective networks, enhancing the prospects of innovation and thus productivity. HSR can act as a major aid to the distribution and sharing of tacit knowledge between clusters that previously could not escape the confinements of their boundaries. This relationship is demonstrated in Figure 1.

Figure 1. A shared prospect between clusters and HSR networks, the movement of knowledge under an expanded knowledge bubble that reaches another regional cluster within 200–800 km (distances based on Hall (1999) and SDG (2004)).

The concept that knowledge leads to innovation, and that clusters and intra-regional knowledge flow enhance innovation, exist in a strong library of literature and have been discussed in brief in Section 2. This enables the understanding of a third key link which is lacking: The bridge between HSR time–space accessibility and inter-regional knowledge flow between clusters, causing greater inter-regional innovation. This paper puts forward a conceptual framework for this bridge (Section 4).

4. Conceptual framework

This review proposes that HSR has a role in the interaction of knowledge carriers across inter-regional clusters. A framework is put forward that builds on aspects of previous knowledge-based literature to formulate a new concept that explains the role of HSR, in the development of an open knowledge economy between clusters.

The development of a regional closed cluster is first assumed to exist in the feedback of its own innovation success and specialisation. This can be simplified into 5 key aspects; the cluster’s infrastructure itself (larger space and infrastructure increase production scale and capacity (Trung et al., 2022)), secondly knowledge accumulation (see Morosini (2004): “knowledge integration between firms as well as institutionalised trust and personal interactions between economic agents are especially strong”). Thirdly the inhabiting knowledge community (a local community sharing knowledge, values and views (see Becattini, 1990)), technology development (shown in Taiwan’s industrial clusters (Chen, 2011)), and lastly innovative output (for example, Falck et al. (2010) found cluster-oriented policies increased the number of innovators). The growth of an industrial cluster’s specialised tacit knowledge allows for innovative capabilities that would lead to productive results (inventions), drawing back in reinvestment and resources, and creating positive feedback. Policy in these closed clusters should favour the learning networks that exist within them to take advantage of the local knowledge economy. However, clusters can be limited in their closed nature, with a lack of knowledge freely flowing among other clusters, hindering efficient innovation.

It is key therefore to first note the synergy which cluster infrastructure and knowledge communities have together when inputting the role of HSR. A knowledge community is restricted in its ability to grow by the infrastructure that they inhabit. At the social knowledge level, a cluster’s infrastructure is argued to be a ceiling to that knowledge expansion, a limit to effective communication (Shibutani, 1955), and knowledge communities will grow very slowly once they have filled that infrastructure space. A knowledge community and the infrastructure of the cluster it exists in, therefore, is a relationship, we define, as the knowledge capacity (KC). KC can be demonstrated in Figure 2, where the community will grow to fill the infrastructure that holds it over time. KC requires both a knowledge community (those who carry knowledge) and the infrastructure to house it efficiently (i.e. a cluster). If a cluster has a policy of infrastructure to allow for innovative potential but a still-developing knowledge community, it will grow and develop only steadily over time. Without the community, the infrastructure will not be utilised and will be an expensive regional development investment with little return. If a KC has a knowledge community but no infrastructure to take advantage of it, there are risks of a region lagging significantly with the restriction of growth and loss of key talent who seek opportunities elsewhere. With this definition, one can begin to explore the resulting impacts of using HSR, to turn multiple closed clusters into open clusters, and provide KC opportunity to grow.

Figure 2. Knowledge capacity of an industrial cluster, defined as the relationship and scale to which a cluster’s knowledge community members exist within the confinements and limitations of the cluster infrastructure provided.

Assuming institutional routines and pre-conditions are met, such as an engaged and interacting knowledge community within an innovative industry, that exists within the industrial structure of the defined cluster, Figure 3 demonstrates the conceptual theoretical framework that expands from Figure 2. Throughout history, there have always existed pathways between regions to share knowledge through social means. An infrastructure bottleneck has restricted the knowledge shared between regional communities. This bottleneck is based on infrastructure that would enable social means of knowledge sharing to exist at its greatest potential. Figure 3 illustrates the development of four different scenarios. Scenario 1 indicates two clusters with their own KC, linked by a narrow bottleneck which would represent two regions that are connected through odd trade and human messengers. This is a more historic form of knowledge exchange during times of lower transport technology, such as before motorised transport, where new knowledge would enter clusters/cities through merchants, travellers and regional messengers by foot or horseback. Therefore, the bottleneck of knowledge transfer reflects the limited amount, frequency, and ability of knowledge to be carried to new locations, and shared between themselves. Upon opening the narrow bottleneck through improvements in transport accessibility between regions, Scenario 2 shows a bottleneck that has been expanded, due to the development and implementation of transport technologies. Examples of Scenario 2 can be shown in general transport improvements throughout history, where separate regions were connected through a more accessible network. For example, the Grand Canal of Sui and Tang Dynasties which improved inter-city development (such as in Kaifeng) through the waterways (Huang et al., 2021) and the railway development of Britain during the 19th centuries Industrial Revolution (for example see Bogart et al., 2017). Without encouraging improved accessibility, regions run the risk of being left behind, with the threat of economically advanced regions diversifying further into high-tech industries while lagging regions exist in a “spatial inequality feedback loop” of low-complex activities (Pinheiro et al., 2022).

Figure 3. The transport technology bottleneck of social knowledge flows between two clusters (A and B) in four scenarios.

More accessible means of transportation between regions increases the rate knowledge can travel socially between regional clusters, through knowledge carriers. However, upon the widening of the bottleneck and thus accessibility between regions, if one region is significantly ahead of the other, there is a risk of brain drain (shown in Scenario 3). Brain drain implies the emigration of knowledgeable individuals to developed regions, creating regional-knowledge imbalances (Agrawal et al., 2011). HSR has been noted to potentially cause brain drain effects in the accumulation of talent in developed cities, such as in China from accumulating talent and labour (Ke et al., 2017; Xie et al., 2021) leading to peripheral cities losing out on enterprise productivity to core cities (Yang et al., 2021) as well as GDP per capita (Yu et al., 2019). Brain drain occurs from a poorer knowledge region to a richer knowledge region when the lacking region does not have the cluster infrastructure to entertain an expansion of KC (Scenario 3). This is also called “tunnel effects” and imperfect competition (Fujita et al., 1999), wherein a lagging region suffers losses with the improvement of transport infrastructure connectivity. These follow the school of thought from New Economic Geography, based on market forces that do not consider the intervention of regional policy which aims to address this uneven development (for example, Scott & Storper, 2003).

With the understanding above, there could be another Scenario 4 (regional balancing). This might be achieved by the balancing of cluster infrastructure between the two connected regions. This framework suggests that knowledge can move from regions with maximised KC to those with larger available cluster infrastructure space through HSR, due to reduced time–space geographies. This would represent the moving, relocation or expansion of firms, research institutes, and establishments of regionally foreign start-ups and academic entities. What is vital is a means of efficient transport that would encourage and allow knowledge to move into a lower KC, demonstrated through a larger bottleneck of knowledge flow. If the cluster infrastructure is being developed, knowledge can move and be shared into a poorer region at a rate that could potentially develop the region and speed up its innovative potential, since the cluster holds the infrastructure to welcome it. Therefore, improving transport widens the bottleneck of inter-regional movement by knowledge carriers. This increase in volume means actors interact and form face-to-face relationships at a high frequency. The bottleneck itself represents the openness of a cluster, at a social knowledge exchange level.

4.1. The HSR bridge

With regional balancing opportunities in place (Scenario 4), Figure 4 demonstrates how HSR, a more efficient form of transport than road and air travel, expands the accessibility bottleneck between two cluster KCs. More frequent face-to-face interactions within a one-day catchment area from one cluster to another give rise to a larger “tacit bubble” with a larger diversity of knowledge carriers between the two localities, as knowledge community members from cluster A will be able to stay and mingle more in cluster B (which previously bore limited access to tacit knowledge) and exploit the infrastructure available to them. As a result, HSR would expand geographical “contact with more skilled neighbours” as Glaeser (1999) urged for the most efficient learning. In a word, the change in tacit knowledge geography induced by the widening of knowledge flow accessibility from HSR connectivity enables the 5 key aspects of a closed cluster’s innovative success to be expanded upon and amplified through enabling relationship building, knowledge exchange, industrial and technological advancement and improved performance across clusters.

Figure 4. The HSR Bridge, the social knowledge landscape of two clusters connected through high-speed rail, amidst a time-space warping of geographies due to significant bottleneck widening in accessibility between the two regions.

Knowledge intermediaries do not necessarily have to be actors of firms, so high-speed transport can serve as an infrastructure intermediary by acting as the bridge within a larger knowledge system (Howells, 2006). The introduction of HSR can trigger innovation through the convergence of knowledge (Bhatt & Kato, 2021), and then maintain significance by acting as a driver for further reactions of knowledge sharing (see Curran (2013) for triggers and drivers in convergence). If HSR infrastructure expands the bottleneck of knowledge flow between two regions, and they hold appropriate cluster infrastructure (enough to hold the expansion of the knowledge community), it may lead to the balancing of lagging regions in their potential development (regional balancing). For example, innovation performance has been shown to increase for cities connected with HSR, due to the promotion of flow of high-quality labour (Bian et al., 2019).

Knowledge convergence driven by HSR can represent the movement of foreign (to the region) high-tech tacit knowledge to local low-tech tacit knowledge, creating the accumulation of knowledge. However, this is at a vaster level as the cluster has been further opened in its knowledge bottleneck. This is as long as the receiving lagging region has the cluster infrastructure in place to receive it, as highlighted by Scenario 4 in Figure 3. This accumulation will be assimilated into a cluster’s innovative potential allowing for a significant increase in previous levels of innovation and thus economic and inventive outputs.

Assimilation will allow for the positive feedback loop of knowledge growth to be larger and more rapid, leading to a low-tech landscape developing into a “developing” high-tech landscape. This follows to what papers described previously, such as; Dong et al. (2020) and Wang and Cai (2020) concerning research, and papers on HSR and knowledge spillovers (Inoue et al., 2017; Lu et al., 2022), exchange (Wang et al., 2022), collaboration (Wang et al., 2022) and productivity (Bhatt & Kato, 2021; Komikado et al., 2021). What has been lacking is a strong conceptual framework towards this line of thinking and a clear explanation of the need for social engagement in HSR research. Frameworks and papers that focus on knowledge flow impacts of transport cannot rely on quantitative measures as evidence, due to the social exchange and nature that is a founding element of the need for travel. Therefore, when it comes to knowledge flows, deeper insights should include social measures and engagements with knowledge carriers and policymakers, to avoid relying on correlative data. Such social exchanges are proposed in this framework to include networking, collaboration, and firm expansions between regions, thanks to the time–space warpage and social knowledge theory, which cannot be measured thorough quantitative means (when linking to HSR impacts).

Qualitative and socially probing methodologies will provide significantly more insight into these types of propositions, than quantitative data, as the deeper interactions and experiences of talent riding on HSR will become more evident, or indeed proven insignificant. Lines of enquires can therefore become established to guide future research that explores the relationship between HSR and industrial clusters when using face-to-face tacit interactions as a key indicator of knowledge flows. For example:

• How can the actors in the knowledge communities be defined, operate, and interact (such as collaborate, build relationships and trust) during the knowledge accumulation process within and across regions?

• To what extent and why do the actors perceive the role of HSR against other modes of transport for creating the enlarged knowledge bubble?

• Transport could be a necessary but not sufficient condition. What limitation might there be? To what extent and how should the role of policymakers or intervention play in making the expanded knowledge bubble work?

5. Concluding remarks and scope for future research

This paper makes two major contributions. First, it has identified a knowledge gap and linked the current seemingly separate literature in HSR and social knowledge theory with cluster economics, highlighting key issues in current works connecting HSR with knowledge flows. Secondly, it has developed a framework to argue the case towards how expanding high-speed transport development, specifically HSR, which reduces time–space geographies, can enable an increase in social knowledge exchange by widening a bottleneck in accessibility that was previously restricting inter-city/region face-to-face interactions. Lines of enquiry for future research on this topic are highlighted and suggest a change in methodological approaches when tacit knowledge flow is analysed.

Regarding the knowledge gap, this paper has shown the depth of theoretical concepts in social knowledge theory, tacit knowledge, and communities of practice. It demonstrates poor bridges between this side of knowledge research and transport, more specifically HSR studies. While studies do explore HSR impacts on knowledge flows, positivist approaches cannot identify inter-personal tacit knowledge exchanges fully. Papers focussing on knowledge indicators such as patents miss an important aspect of actor face-to-face interactions. For example, co-patents involve not only collaborations and exchanges but also development, funding/partnerships through various means for industry growth. Similarly, to understand knowledge exchange as a result of HSR, one cannot rely on economic indicators and database analyses to comprehend the movement of individual knowledge carriers, and how/why actors collaborate and exchange knowledge. A recent review paper on the socio-economic impacts of HSR published in this journal (Cheng & Chen, 2022) reflects the limitation of deeper understanding through predominantly quantitative modelling methods. In a similar perspective, Ludwig et al. (2024) explain a widening theory-practice gap forming, arguing a need for more philosophical understandings within urban planning research, with an intensification of positivist and objective pursuits leading to subjective knowledge becoming difficult to measure.

This review has contributed to the realm of tacit knowledge exchange between regions, bridging it with HSR studies. It has highlighted how tacit knowledge thrives in clusters, and how HSR can utilise this model to create a larger inter-regional metropolis of knowledge that exists within a one-day catchment bubble. This provides noteworthy paths towards future research and planning studies as geographies may begin to be redefined within the warped proximity effect of HSR accessibility. The framework constructed highlights how HSR connectivity can increase the ability to share social (tacit) knowledge between two regions now connected. By providing new accessibility means, more individuals are able to travel within a shorter time, increasing face-to-face contact and relationship building among actors. This leads towards an increase in regional innovation and knowledge sharing. As a result, two regions may enter into one larger metropolis knowledge bubble. This framework is novel within HSR literature, as it provides clarity and linkages between cluster and localised knowledge theories (i.e. Bathelt et al., 2004; Malmberg & Maskell, 2002; Porter, 2001) with the time–space warping/convergence accessibility effects of HSR (i.e. Jiao et al., 2014; Spiekermann & Wegener, 1994). It has demonstrated how HSR can widen the bottleneck between regions in a way that allows for geographies of tacit knowledge boundaries to be potentially redefined, and to be researched further. The framework offers a counter to “brain drain”, suggesting a plausible movement of knowledge to a region of less development if a cluster’s infrastructure is developed enough to house talent.

HSR studies on knowledge exchange, where face-to-face engagements are demonstrated, must orientate towards the social aspect of actors. This requires a methodology targeted towards process tracing, to examine and evaluate knowledge flow based on reading reports and papers, as well as interpersonal interpretations from in-depth discussions with actors at the forefront of knowledge exchange. More qualitative-based studies, while time-consuming, should be encouraged and constructed to analyse experiences, opinions and community engagements that are often lost in quantitative approaches. Particularly, in-person methodologies amongst a greater selection of case studies (the majority presently in China) would benefit from a focus towards picking up tacit hints in interactions. This paper therefore starts to propose lines of enquiry that can be adapted or developed into future studies to investigate social interactions at the tacit dimension from HSR ridership and talent mobility.

This review encourages further research that examines multiple questions and methodologies. Questions should start to ask whether and how HSR can indeed create a social knowledge bubble between multiple clusters. If a knowledge bubble is formed, it can redefine regional clusters, inter-regional geographies, and transport relations to inter-regional innovative potential. Further, how could the boundaries of two clusters after HSR inter-regional connectivity be defined within a one-day catchment area? With increased HSR accessibility, are connected clusters singular localities or multiple? Does HSR reinvent spatial proximities within innovative industries? And to what extent does HSR offer opportunities for areas that lack a strong knowledge community to benefit from other more developed regions (such as in high-tech industries)? Future research should start to move away from database work, using patents, publications, and economic indicators, as main indicators for tacit knowledge exchanges. While providing insight, such approaches lack in-depth evidence of exchanges. Social studies are better suited for representing social knowledge flows from knowledge carrier interactions and experiences. Future work in transport and HSR research that focuses on tacit knowledge flows should therefore explore interpersonal qualitative analysis primarily.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the University of Liverpool (UoL) and National Tsing-Hua University (NTHU) under the UoL/NTHU Dual-PhD Programme Scholarship.