Territorial servitisation in Italy: Towards territorial recoupling

Abstract

Integration between manufacturing and service activities currently represents a core source of competitiveness for firms and territories. Furthermore, a growing literature analyses the phenomenon of territorial servitisation, namely the co-location of manufacturing and Knowledge Intensive Business Services (KIBS) activities. This article investigates the process of territorial servitisation in Italy by analysing the diffusion of KIBS activities in four different types of local production systems. It distinguishes between professional KIBS and technological KIBS. The magnitude and growth of KIBS and their co-location with manufacturing industries are described by means of firm employment data, location quotients, co-location maps and econometric analysis. Results show that Italy is still experimenting a generalised decoupling phenomenon between manufacturing and KIBS activities, especially professional KIBS, with the latter concentrated in large urban local systems. Technological KIBS also tend to decouple from manufacturing local production systems, especially from those dominated by large enterprises and those specialised in low-tech sectors. Despite the revealed spatial decoupling, we find an ongoing process of coupling, that is, closer proximity, between more traditional manufacturing industries and both professional and technological KIBS.

Keywords

Co-location Italian local production systems KIBS manufacturing territorial servitisation

Introduction

Digital and green transitions (De Propris and Bailey, 2020) are redrawing the process of value creation in manufacturing industries with systemic repercussions for the local production systems (LPSs) where firms’ individual decisions are taken. Firms face technological pressures and new market demands; at the same time, value chains are being re-organised with old competences unlearnt (Bellandi et al., 2021), while new ones are being acquired. The opportunities offered by 4.0 technologies (Capello and Lenzi, 2021) rest on the ability of places and industries to embrace the transformative and pervasive change that is required.

This article focuses on a specific aspect related to the diffusion of 4.0 technologies that sees the more complex and synergetic integration of manufacturing activities with services as required to address or create markets that are more polarised, fast-moving – with short product life cycles – and dominated by a continuum of loops involving innovation-production-consumption (De Propris and Bailey, 2020). Emergent business models rest on processes of servitisation (Vandermerwe and Rada, 1988) that affect in particular manufacturing activities (Dimache and Roche, 2013), critical for the competitiveness of manufacturing firms and territories. A servitised manufacturing sector requires manufacturing firms in LPSs to access new competences and knowledge related to 4.0 technologies to design new strategies for value creation and identify new pathways to competitiveness. Manufacturing firms can either develop the required new competences in-house, or they can procure highly knowledge-intensive and specialised services from expert service firms; the latter are referred to as KIBS firms. Reliance on external providers depends on several factors, including awareness, needs, proximity and firm size (Brittes Benitez et al., 2020). A growing literature has analysed the phenomenon of territorial servitisation (Vendrell-Herrero and Wilson, 2017), namely the co-location of manufacturing firms and KIBS, and the advantages that this offers in terms of service specialisation, awareness of and access to new knowledge, co-production and co-innovation and productivity (Lafuente et al., 2017; Lombardi et al., 2022; Sforzi and Boix, 2019). Although the current literature has looked at the evolutionary nature of the servitisation process and at the geography of servitisation, more research is needed to explore the territorial dynamics and synergies between manufacturing sectors and KIBS at the local level.

The aim of the article is to shed light on the forms of servitisation across different types of LPSs in Italy and identify processes of territorial servitisation. We will address two research questions (RQs). RQ1 explores whether manufacturing territorial servitisation varies in different types of LPSs and depending on the territorial distribution of two different types of KIBS: professional KIBS (P-KIBS) and technological KIBS (T-KIBS). RQ2 asks whether the evolution of P-KIBS and T-KIBS and manufacturing sectors in LPSs is evidence of a coupling or decoupling process over time; the former would suggest an ongoing process of territorial servitisation across Italian LPSs.

The novelties of the article, therefore, are threefold. First, by addressing the gap in the current literature on territorial specialisation, it sheds light on the presence of co-location between KIBS and manufacturing activities by looking at sector composition within individual LPSs but also at the sector specialisations of contiguous places. This allows the identification of areas of spatial manufacturing continuity as well as areas where manufacturing activities neighbour KIBS activities. Second, the article distinguishes T-KIBS from P-KIBS (Baldoni et al., 2022; Vaillant et al., 2021) and is therefore able to flag up the co-location of manufacturing activities with services more strictly related with 4.0 technologies. We specifically distinguish high-tech from low-tech manufacturing industries. Finally, the article presents static and dynamic analyses through econometric tools that deliver statistical evidence of the current interplay between manufacturing and KIBS in the different types of LPSs, and whether this is part of a process of coupling over time and therefore of territorial servitisation evolution.

The analysis focuses on Italian LPSs. Empirically, these are operationalised using Local Labour Systems (LLSs), which are functional economic areas delineated by commuting flows and defined by the Italian National Institute of Statistics (ISTAT). For clarity and consistency with the literature on territorial servitisation, we refer to these units as LPSs throughout the article. The dataset includes employment by sector and municipality, which we re-aggregate at the LPS level, and computes location quotients (LQs) for the period 2012–2018.

The article proceeds as follows: the Knowledge, KIBS and territorial servitisation section provides the conceptual framework of the article by drawing on the literature on the knowledge economy and tracking the changing role of services in the production system with a discussion on territorial servitisation. The Methodology section presents the data and the methdological approaches, and the Results section discusses the findings. Some concluding remarks complete the article.

Knowledge, KIBS and territorial servitisation

The role of services in the economy has evolved over time. They have always played an important role especially in urban economies in the form of retailing, public services (including academia), entertainment and hospitality (Chapain et al., 2010; Florida, 2002; Glaeser, 1998). In particular, the literature distinguishes between knowledge-intensive services and less knowledge-intensive services (see Appendix 1, Table 6) (Aslesen and Isaksen, 2007; Capello et al., 2012; Lafuente et al., 2017). A turn in our understanding of the much more synergetic role that some services can have especially with manufacturing was discussed in the early 2000s in the literature on the knowledge economy (Cooke, 2002) and the dematerialisation of value creation in manufacturing (Cooke et al., 2007; OECD, 2004; Rooney et al., 2005). The diffusion of communications technologies (i.e. Internet, personal computers) in the 1990s introduced an additional resource to labour and capital in production, in the form of knowledge, allowing a shift from a resource-based economy to a knowledge-based economy (Houghton and Sheehan, 2000). Knowledge became to be understood as the main driver of innovation and, therefore, so was value creation throughout the production process from design to production and further down the value chain to marketing and distribution. In particular, there was a sub-set of the service sector that had a critical role to play in adding value to manufacturing.

KIBS (see Appendix 1, Table 7) indeed became instrumental for manufacturing industries in providing and channelling specific service-related competencies that created value in a knowledge-based economy (Lafuente et al., 2019). In imperfectly competitive global markets where firms’ competitive advantage was centred around product differentiation, KIBS were able to contribute to firms’ strategies towards globalisation (exporting, storage, packaging), product differentiation (design, innovation and creativity; marketing; branding; training, advertising, technical) and firms’ consolidation (organisational change, legal and financial) (Sforzi and Boix, 2019). Such highly value-adding services were found to be critical for the competitiveness of manufacturing firms in advanced economies as the constituent knowledge provided intangible resources for firms’ innovation capabilities. In particular, the importance of KIBS was first recognised in topping and tailing the production process as described in the smile curve (Shih, 1996).

This literature looked at the interplay between services and manufacturing from a supply chain perspective which is fundamentally trans-spatial in the light of the debate on the decoupling between urban and sub-urban spaces in terms of specialisation. The abundance of creative talent, excellent connectivity (transport and communications), Jacobian economies derived from the rich mix of economic activities (Glaeser, 2000) and access to amenities (Florida, 2002) made cities beacons of the knowledge economy, as against manufacturing sub-urban areas. This geographical decoupling could also be scaled up in the so-defined second unbundling (Baldwin, 2016) with highly value-adding services in advanced economies and labour-intensive manufacturing activities in emerging economies (see the literature on global value chains, Gereffi et al., 2005).

The role of services as highly specialised intangible competences that can transform the nature of the manufacturing process and product has been captured by the fast-moving literature on servitisation. The term servitisation was introduced to describe the change in the strategy and capabilities of manufacturing firms from selling products to selling the use of the same good (Baines et al., 2007, 2009; Vandermerwe and Rada, 1988). Servitisation, therefore, underpins a new manufacturing model which is characterised by a high intensity product–service innovation strategy, as well as fast technological adoption and upgrading together with shorter product life cycles inviting closer exchanges between ‘the making of things’ and the services attached to that. Servitised manufacturing (Dimache and Roche, 2013) coincides with the contamination of service practises and strategies to manufacturing (Baines et al., 2017).

The literature on servitisation has developed in many novel directions as evidence of the fundamental and structural change in the nature of services as a pivotal pool of value-adding competencies able to trigger process and product innovation. One line of research focuses on understanding the nature of a new model of servitised manufacturing (Bustinza et al., 2022; Lafuente et al., 2023), which sees servitisation as a product-service innovation process that deploys value-adding knowledge-based services along the production process and, in a continuum, to users and customers (Araya et al., 2020; Marino and Trapasso, 2020). In particular, the current debate links servitisation with digital technologies associated with the Fourth Industrial Revolution (De Propris and Bellandi, 2021; Opazo-Basáez et al., 2024; Vaillant et al., 2021), marking a broader shift towards a smart manufacturing model characterised by the penetration of digital technologies throughout the lifetime of products (Bustinza et al., 2022). Vendrell-Herrero et al. (2024) identify two forms of digital servitisation: one driven by standardisation priorities, and the other by customisation strategies. This distinction can be crucial when servitisation is adopted in manufacturing industries with different levels of technological vs design intensity.

Smart manufacturing is often referred to in relation to industries reliant on cyber-physical factory systems and digitally connected supply chains, and with business models based on replicability and scalability of operations, cost-efficiency and adaptability. However, a second line of research in the servitisation literature looks at it as a process of organisational change, and this allows for a more dynamic and evolutionary understanding of the product-service innovation process (Baines et al., 2020). The journey that firms undergo to servitise is found to depend on market as well as on internal factors (Baines et al., 2020) coupled with a strong technological push. Whether internal or external (Vendrell-Herrero and Wilson, 2017; Xing et al., 2023), the digital servitisation process rests on access to knowledge that can be distant from the one characterising the firm. In this respect, design-intensive industries are more likely to externalise servitisation with relational or system solutions (Hauknes and Knell, 2009; Xing et al., 2023).

The third line of research studies the role of space and place in the servitisation process of hybrid value chains (Lafuente et al., 2019), namely it explores whether the location and agglomeration of KIBS in certain places depend on related (other KIBS) or unrelated (specifically manufacturing) industries. It also explores if there is a converging or diverging spatial pattern between KIBS and manufacturing industries. The concepts of territorial servitisation (De Propris and Storai, 2019; Lafuente et al., 2017), place-based servitisation (Sforzi and Boix, 2019) and recently of product-service innovation ecosystem (Lafuente et al., 2023; Vaillant et al., 2021) capture a new understanding of the interconnections and synergies that can emerge from the co-location of manufacturing activities and KIBS. Because of the complementary nature of their knowledge bases and the dovetailed close relationships between KIBS, one could see why co-location brings about evident benefits in terms of agglomeration, external economies, specialisation and knowledge spillovers.

Empirical studies on territorial servitisation highlight its diverse manifestations and impacts across different industries and regions. The arguments have started to appreciate that it was not agglomeration per se that benefitted firms, but access to collaborative networks and knowledge spillovers (see, for instance, Albors-Garrigos et al., 2012, on the role of KIBS in fostering innovation and entrepreneurship in urban regions). Special attention has been paid to the interconnections between manufacturing firms and KIBS with an emphasis on the hybrid nature of the local supply chains as this allows KIBS to interject throughout the various phases of the production processes (Araya et al., 2020; Lafuente et al., 2019, 2023). Alternatively, KIBS could enable value-adding fit, with KIBS contributing with idiosyncratic knowledge to the local manufacturing specialisation (Miles and Snow, 1984) or could trigger successful processes of servitisation in firms because of proximity (Opazo-Basáez et al., 2020; Vaillant et al., 2021). Bustinza et al. (2022) argue that territorial servitisation in the form of a localised product-service innovation system is critical to equip firms with the relevant technology to add value to their manufacturing processes, and in particular, smart manufacturing, in other words, proximity enhances knowledge sharing on the compatibility of smart technology depending on the manufacturing specialisation. This is reiterated by Vendrell-Herrero et al. (2024) when comparing customisation and standardisation servitisation strategies. Focusing on industrial districts as a specific case of manufacturing production systems dominated by small- and medium-sized enterprises (Becattini et al., 2009), Baldoni et al. (2022) find, however, that in Italy, industrial districts struggle to make such transition with KIBS competences still spatially decoupled from those districts especially specialised in low-tech manufacturing activities.

Although the existing literature on servitisation has looked at aspects that relate to the evolutionary nature of the process of servitisation on one side and to the geography of servitisation activities on the other, research is still needed to explore to what extent the territorial dynamics between manufacturing and KIBS depend on the different types of LPS and the nature of KIBS specialisation; what forces might drive territorial servitisation and also how they might evolve over time. The article addresses these issues, first, by distinguishing four types of LPSs depending on their sector specialisation and firms’ size – small or large – (Sforzi and Boix, 2019) and on population density to pull out the urban effect (Glaeser, 2000). This analysis allows the ability to shed light on the geographical distribution of KIBS in different types of LPSs and, therefore, if there is evidence of territorial servitisation in some rather than other LPSs. The article also distinguishes P-KIBS and T-KIBS (see Appendix 1, Table 7), as well as low-tech and high-tech manufacturing activities (see Appendix 1, Table 8), to allow for a more fine-grained analysis that takes into account the specific sector specialisms of a local system and to uncover how such specialisms may impact the coupling or decoupling processes. This first RQ is therefore:

RQ1: Does the territorial distribution of KIBS activities (specifically P-KIBS and T-KIBS) in different types of LPSs point to evidence of manufacturing territorial servitisation? Does the servitisation process vary across different types of LPSs?

Second, from a dynamic perspective, the analysis moves on to trace the evolution of location processes of KIBS especially in manufacturing local systems. It investigates whether the spatial dynamics of KIBS industries point to coupling or decoupling processes, depending on whether KIBS presence in manufacturing systems shows signs of growing or remaining stable. A coupling trend would suggest a process in motion of territorial servitisation, whereby KIBS are attracted to locate nearby manufacturing activities driven by an expanding demand by manufacturing firms for specialised services. Hence, the second RQ is:

RQ2: Does the evolution of KIBS and manufacturing over time point to evidence of a coupling or decoupling process? In other words, is there an ongoing process of territorial servitisation across Italian LPSs?

Methodology

The empirical analysis relies on employment data from the Italian Statistical Archive of Active Firms (ASIA) provided by ISTAT, combined with ISTAT’s official identification¹ and classification of LLS. These latter are functional economic areas delineated by commuting patterns and are commonly used as proxies for LPSs (De Propris, 2005). In this article, we refer to these units as LPS to reflect our theoretical focus, while acknowledging that they are empirically constructed using LLS boundaries.

The analysis covers the period 2012–2018. We extract employment data disaggregated by municipality and two-digit sector codes and re-aggregate it at the LPS level. Sectors are grouped into five main aggregates: high-tech manufacturing (HT-MAN), low-tech manufacturing (LT-MAN),² knowledge-intensive services (KIS), less knowledge-intensive services (LKIS) and KIBS.³ A residual category is created to group all remaining sectors. We use the EUROSTAT classification of ‘High-tech industry and knowledge-intensive services’ (see Appendix 1, Tables 6 and 8), whereas the KIBS aggregate is defined according to Vaillant et al. (2021) and spilt into T-KIBS and P-KIBS (see Appendix 1, Table 7). This distinction allows us to highlight service activities linked to Industry 4.0 technologies and their interaction with manufacturing firms of varying technological intensity.

We follow ISTAT’s (2014, 2015) taxonomy and classify LPSs into the following four types: LPSs in large urban areas (LUAs)⁴; manufacturing LPSs with a prevalence of Small and Medium Enterprises (SMEs); manufacturing LPSs with a prevalence of large firms and a residual group labelled ‘Other territories.’ LUAs are of particular interest due to their centrality in communication and information networks and their concentration of advanced service activities and organisational competences that foster KIBS development (Glaeser, 2000). In contrast, manufacturing-specialised LPSs host transformation activities that could benefit from interacting with KIBS if adopting servitisation strategies. We distinguish manufacturing LPSs by the dominant firm size because firm size influences a company’s capacity to access and absorb KIBS competences (Brunow et al., 2020). Large firms may develop such competences in-house, whereas small firms often rely on specialised external providers (Bustinza et al., 2022). This distinction also reflects the uniqueness of the Italian manufacturing sector, which is characterised by a large share of small- and medium-sized enterprises (Vrontis et al., 2018). Employment data is used to compute LQs⁵ (De Propris, 2005) and growth differentials by sector and LPS. LQs help identify local specialisations, map sectoral agglomeration and detect co-location dynamics between KIBS and manufacturing.

To address RQ1, we examine co-location patterns both within and across LPSs. Within each LPS type, we contrast the average LQs of KIBS (T-KIBS and P-KIBS) and manufacturing activities (high tech and low tech). To assess co-location across LPSs, we develop bivariate Local Indicators of Spatial Association (LISA) cluster maps based on the local Moran’s I statistic (Anselin, 2019).⁶ These maps reveal the spatial relationships between the specialisation of each LPS and that of its neighbours, allowing for a classification of LPSs based on their own KIBS specialisation and that of surrounding areas. To address RQ2, we examine the growth of employment and the evolution of LQs in the period across LPSs.

We complement the descriptive analyses with a dynamic spatial panel data model with two-way fixed effects. This model captures the temporal persistence in KIBS specialisation and spatial dependence across LPSs, reflecting the inherent dynamic and interdependent nature of regional economic systems (Fischer, 2021). By incorporating fixed effects, the model also addresses potential biases arising from unobserved heterogeneity. Such modelling approaches are widely used in regional economics to examine dynamics in employment, innovation and productivity and to estimate policy impacts (Billé et al., 2023; Bouayad-Agha et al., 2013; Elhorst et al., 2013; Fingleton, 2024; Halleck Vega and Elhorst, 2013; Rios, 2017; Wanzenböck and Piribauer, 2016). The specification used is as follows:

\begin{array}{l} L Q_{i t, n}^{K I B S} = μ_{i} + φ L Q_{i t - 1, n}^{K I B S} + ρ W^{(1)} L Q_{i t, n}^{K I B S} \\ + Σ α_{m} W^{(2)} L Q_{i t, m}^{M A N} \\ + Σ γ_{k, m} T Y P E_{i, k} L Q_{i t, m}^{M A N} \\ + Σ β_{k, m} T Y P E_{i, k} L Q_{i t, m}^{M A N} T R E N D_{t} \\ + d_{t} + δ {\bar{L Q}}_{i t}^{O T H E R} + ε_{i t, n} \end{array}

(1)

where $L Q_{i t, n}^{K I B S}$ represents the LQ KIBS of type $n$ ( $n$ = KIBS, T-KIBS, P-KIBS) in LPS $i$ at time $t$ . The terms $μ_{i}$ and $d_{t}$ denote the LPS and year fixed-effects, respectively. The coefficient $φ$ captures temporal persistence, through the lagged dependent variable, while $ρ$ and $α_{m}$ capture spatial dependence, via two spatial weight matrices⁷: $W^{(1)}$ and $W^{(2)}$ . The term $L Q_{i t, m}^{M A N}$ refers to the LQ for manufacturing of type $m$ ( $m$ = HT-MAN, LT-MAN), and ${\bar{L Q}}_{i t}^{O T H E R}$ is the average LQ of other economic activities (KIS, LKIS and other activities). The coefficients $γ_{k, m}$ represent the spatial coupling between manufacturing and KIBS across LPS types, based on the interaction terms between manufacturing specialisation and the dummy for the type of LPS ( $T Y P E_{i, k}$ ). A significant negative coefficient would signal a spatial decoupling process and provide a statistical test for RQ1.

To test statistically RQ2, we focus on the sets of coefficients $β_{k, m}$ . These capture temporal recoupling trends, derived from the interaction of manufacturing specialisation, LPS type and a linear time trend ( $T R E N D_{t}$ ). A positive coefficient for these interaction terms would signal the existence of an ongoing coupling process.⁸ The model is estimated using the Bias Corrected Least Squares Dummy Variable (BCLSDV) estimator (Baldoni, 2023; Elhorst, 2010; Yu et al., 2008).

Results

Territorial decoupling between manufacturing and KIBS (RQ1)

This section discusses the results that address RQ1. The analysis combines descriptive statistics with econometric modelling that tests whether there is evidence of territorial decoupling between manufacturing and KIBS – specifically P-KIBS and T-KIBS – across different types of LPS.

Descriptive analysis of spatial distribution

The data shows that half of KIBS employment is in LPSs in LUAs, about 20% is in manufacturing LPSs of SMEs, followed by manufacturing LPSs of large firms (LPS-LE). Approximately another 20% of KIBS employment is in mixed LPSs, referred to as ‘other territories’. Specifically, both P-KIBS and T-KIBS are found to be more present in SME-based manufacturing LPSs than in those dominated by large enterprises (Table 1).

Table 1.

Total, manufacturing and KIBS employment (2018).

	Total		Manufacturing				KIBS
	Total		HT-MAN		LT-MAN		P-KIBS		T-KIBS
LPS in large urban areas	5,862,435	33.9%	322,734	28.4%	507,713	19.6%	785,107	43.5%	314,500	57.6%
Manufacturing LPS of SMEs	4,567,165	26.4%	411,847	36.3%	1,107,714	42.8%	413,094	22.9%	87,301	16.0%
Manufacturing LPS of LEs	2,058,724	11.9%	237,947	20.9%	421,000	16.2%	185,284	10.2%	40,222	7.3%
Other territories	4,799,566	27.7%	162,164	14.2%	551,382	21.3%	419,578	23.2%	103,588	18.9%
Total	17,287,890	100.0%	1,134,691	100.0%	2,587,808	100.0%	1,803,064	100.0%	545,611	100.0%

Source: Authors elaboration on ASIA and ISTAT data.

Breakdown by LPS typology.

With respect to manufacturing activities, the evidence confirms that the bulk of employment remains concentrated in LPSs of SMEs, followed by LUAs and then LPSs of large enterprises. This is especially true for low-tech manufacturing activities, whose employment share in LPSs of SMEs is 42.8%, compared to 36.3% for high-tech manufacturing activities. Interestingly, we found that LUAs present almost one third of the total employment in high-tech manufacturing. This is consistent with the morphology of Italian industries characterised by SMEs which tend to be geographically spread in sub-urban spaces, but with a not negligible presence in some urban areas.

The geographical concentration of economic activities is further illustrated using location quotients. These are presented visually through Figures 1 and 2. The data shows that P-KIBS are relatively dispersed but present high concentrations in both manufacturing-intensive and urban areas such as Bologna, Milan, Genoa, Turin, Rome and Florence. T-KIBS are more spatially concentrated in northern and central urban areas. In contrast, manufacturing activities – particularly low-tech – are primarily located outside of LUAs, particularly in the North (e.g. Piedmont, Veneto, Emilia Romagna, Toscana and Marche). Still, some urban areas in the North – such as Turin and Bologna – present a high concentration of high-tech manufacturing activities.

Figure 1.

LQs KIBS (a), T-KIBS (b) and P-KIBS (c) across LPSs.

Figure 2.

LQs of HT-MAN (a) and LT-MAN (b) across LPS.

Average LQs of 2018 (Table 2) indicate that KIBS (including P-KIBS and T-KIBS), KIS and LKIS are concentrated in LUAs, while high-tech manufacturing is concentrated in LPSs of large enterprises. Low-tech manufacturing shows high concentration in LPSs of SMEs instead. These patterns seem to indicate a strong spatial decoupling: KIBS activities are generally urban, while manufacturing activities are more often found in non-urban LPSs. This reinforces a well-known size and technology divide in the Italian industrial landscape.

Table 2.

Average LQ – 2018.

LPS typology	KIS	KIBS	T-KIBS	P-KIBS	LKIS	HT-MAN	LT-MAN	Other sectors
LPS in large urban areas	1.144	1.247	1.336	1.220	1.063	0.809	0.634	0.892
Manufacturing LPS of SMEs	0.726	0.681	0.464	0.746	0.843	1.220	1.808	1.081
Manufacturing LPS of LEs	0.812	0.685	0.462	0.753	0.840	1.771	1.501	1.099
Other territories	1.014	0.608	0.413	0.668	1.211	0.291	0.788	1.380

Source: Authors elaboration with ASIA and ISTAT data.

Co-location between manufacturing activities and KIBS in contiguous LPSs

As defined, LPSs capture the spatial distribution of economic activities at a very fine-grained level. However, what happens around them is to some extent equally relevant to understand the trajectories of change for two reasons. First, proximity with an urban centre (especially large urban centres) might allow access to knowledge that does not need to be developed either internally to the firm or to the local system. Second, the contiguity between LPSs specialised in either manufacturing or KIBS might trigger cross-sector synergies leading to new variants of territorial servitisation.

To better understand whether manufacturing and services locate in neighbouring areas, we investigate co-location patterns between KIBS and manufacturing activities in contiguous LPSs using 2018 data. Figure 3 presents bivariate LISA cluster maps based on LQs for KIBS and high- and low-tech manufacturing. For instance, in Figure 3(a), pink areas indicate LPSs with high KIBS LQs neighbouring LPSs with low HT-MAN LQs.

Figure 3.

2018 – Co-location maps of LQ KIBS and HT-MAN (a) and LT-MAN (b).

The maps suggest that, especially in northern Italy, urban centres are surrounded by manufacturing LPSs, forming potential functional clusters. For example, around Milan and Bologna, we observe high KIBS concentrations adjacent to high-tech manufacturing LPSs. In the South, urban centres often show high KIBS concentrations surrounded by low or no manufacturing specialisation. When disaggregating by KIBS and manufacturing type, similar patterns emerge. T-KIBS in urban centres co-locate with neighbouring LPSs specialised in high-tech manufacturing and low-tech manufacturing in the North (Figure 4). P-KIBS also show stronger spatial contiguity with low-tech manufacturing in central-eastern regions (Figure 5). These patterns support the notion of proximity beyond LPS borders, implying potential for knowledge spillovers even in the absence of direct co-location.

Figure 4.

2018 – Co-location maps of T-KIBS and HT-MAN (a) and LT-MAN (b).

Figure 5.

2018 – Co-location maps of P-KIBS and HT-MAN (a) and LT-MAN (b).

Econometric evidence of decoupling

To confirm our findings, we estimate a dynamic spatial panel model. Table 3 presents model results for aggregate KIBS (M1), T-KIBS (M2) and P-KIBS (M3).

Table 3.

BCLSDV estimates of models for aggregates LQ KIBS (M1), LQ T-KIBS (M2) and LQ P-KIBS (M3).

		M1	M2	M3
Variable	Coefficient	LQ KIBS	LQ T-KIBS	LQ P-KIBS
LQ KIBS (t-1)	ϕ	0.546 (0.014)***	0.604 (0.015)***	0.581 (0.015)***
W⁽¹⁾ – LQ KIBS	ρ	0.148 (0.045)***	-0.090 (0.101)	0.154 (0.044)***
$W^{(2)}$ – LQ HT-MAN	$α_{m}$	-0.090 (0.041)**	0.245 (0.097)**	-0.182 (0.056)***
$W^{(2)}$ – LQ LT-MAN	$α_{m}$	-0.095 (0.057)*	-0.217 (0.134)	-0.090 (0.078)
TYPE (LE) – LQ HT-MAN	$γ_{k, m}$	-0.160 (0.013)***	-0.072 (0.026)***	-0.186 (0.016)***
TYPE (LE) – LQ LT-MAN		-0.266 (0.036)***	-0.127 (0.073)*	-0.296 (0.045)***
TYPE (LUA) – LQ HT-MAN		-0.077 (0.182)	-0.416 (0.371)	0.018 (0.226)
TYPE (LUA) – LQ LT-MAN		-0.218 (0.397)	-0.483 (0.808)	-0.196 (0.492)
TYPE (Other) – LQ HT-MAN		-0.117 (0.015)***	0.006 (0.031)	-0.153 (0.019)***
TYPE (Other) – LQ LT-MAN		-0.184 (0.018)***	-0.117 (0.037)***	-0.200 (0.022)***
TYPE (SME) – LQ HT-MAN		-0.129 (0.014)***	-0.030 (0.028)	-0.154 (0.017)***
TYPE (SME) – LQ LT-MAN		-0.303 (0.024)***	-0.176 (0.048)***	-0.334 (0.029)***
TYPE (LE) – LQ HT-MAN – TREND	$β_{k, m}$	0.003 (0.001)***	0.002 (0.002)	0.003 (0.001)***
TYPE (LE) – LQ LT-MAN – TREND		0.009 (0.001)***	0.004 (0.002)*	0.01 (0.001)***
TYPE (LUA) – LQ HT-MAN – TREND		0.001 (0.006)	-0.002 (0.012)	0.002 (0.008)
TYPE (LUA) – LQ LT-MAN – TREND		0.008 (0.009)	-0.006 (0.019)	0.014 (0.012)
TYPE (Other) – LQ HT-MAN – TREND		0.007 (0.001)***	-0.005 (0.003)*	0.011 (0.002)***
TYPE (Other) – LQ LT-MAN – TREND		0.006 (0.001)***	0.007 (0.003)**	0.006 (0.002)***
TYPE (SME) – LQ HT-MAN – TREND		0.005 (0.001)***	-0.001 (0.002)	0.006 (0.001)***
TYPE (SME) – LQ LT-MAN – TREND		0.006 (0.001)***	0.004 (0.002)**	0.007 (0.001)***
AVG LQ Other	$δ$	-0.489 (0.023)***	-0.376 (0.046)***	-0.511 (0.029)***
Time dummies	$d_{t}$	Yes	Yes	Yes

Notes: Standard errors in parentheses. $W^{(1)}$ is based on a radius of 150 km (M1), 190 km (M2) and 140 km (M3). $W^{(2)}$ is based on a radius of 140 km (M1), 200 km (M2) and 180 km (M3).

Statistical significance: ***p < 0.01, **p < 0.05, *p < 0.1.

The autoregressive coefficients are highly significant across all models, indicating strong temporal persistence in KIBS specialisation. The spatial lag for LQ-KIBS is positive and significant in M1 and M3 but not in M2, suggesting that T-KIBS do not tend to cluster in space as P-KIBS.

The relation between LQ KIBS and the spatial lag of manufacturing (both high and low tech) depends on the type of KIBS considered. This relation is positive and highly significant between T-KIBS and high-tech manufacturing, whereas it is found to be either negative or non-significant for aggregate KIBS and for P-KIBS. Evidence shows therefore that LPSs specialised in T-KIBS seem to act as gravity centres for the surrounding high-tech manufacturing activities, and thus, one key finding of the analysis is that co-location patterns need to be seen to extend beyond the borders of LPSs, to include contiguous ones, in a continuum of complementary and interconnected economic sector activities.

In the model, the interaction terms between manufacturing LQs and LPS types confirm the decoupling observed in the descriptive analysis. This decoupling appears to be widespread for P-KIBS and manufacturing activities, with the exception of LUAs. A similar pattern is found between T-KIBS and low-tech manufacturing across all LPS except LUAs, and between T-KIBS and high-tech manufacturing, but only in LPSs of large enterprises. These findings support the notion that T-KIBS tend to agglomerate in cities and do not follow manufacturing specialisations – especially the low-tech one – in peripheral areas. This may also reflect the fact that large firms in LPS of large enterprises are more capable of developing advanced service competences internally, reducing their dependence on local KIBS providers.

Evolution towards recoupling (RQ2)

This section discusses whether there is evidence of dynamic recoupling, in the form of a process of spatial co-location between KIBS and manufacturing over time (RQ2).

Dynamic descriptive trends

Employment growth trends (Table 4) show that KIBS grew rapidly between 2012 and 2018 across all types of LPS, but most notably in those manufacturing ones dominated by SMEs and large enterprises. In particular, P-KIBS grew by over 24% in LPS of SMEs, and T-KIBS grew at a rate nearly twice the national average. This growth surpasses that observed in LUAs, indicating a geographical redistribution of KIBS employment.

Table 4.

Employment growth rates: 2012–2018.

LPS typology	KIS	KIBS	T-KIBS	P-KIBS	LKIS	HT-MAN	LT-MAN	Other sectors
LPS in large urban areas	10.66%	16.44%	6.77%	20.82%	9.62%	-6.41%	-1.81%	-7.62%
Manufacturing LPS of SMEs	8.64%	22.92%	15.45%	24.62%	3.56%	3.63%	-2.94%	-15.28%
Manufacturing LPS of LEs	8.62%	25.27%	17.56%	27.07%	2.40%	-1.64%	-6.87%	-15.39%
Other territories	9.15%	12.30%	7.45%	13.57%	3.31%	2.31%	-4.60%	-14.14%
National average	9.58%	17.59%	8.95%	20.48%	5.41%	-0.70%	-3.74%	-12.84%

Source: Authors elaboration with ASIA and ISTAT data.

LQ changes over time show an increase in KIBS specialisation in LPS of SMEs and LPS of large enterprises, while LUAs experienced a slight decline (see Table 5). This pattern is reinforced by increasing LQs for both high-tech manufacturing and low-tech manufacturing in those same LPSs. These trends suggest that although the current distribution remains skewed, a recoupling process seems to be underway.

Table 5.

LQs variation: 2012–2018.

LPS typology	KIS	KIBS	T-KIBS	P-KIBS	LKIS	HT-MAN	LT-MAN	Other sectors
LPS in large urban areas	-0.01	-0.04	-0.09	-0.01	-0.01	-0.05	-0.01	0.02
Manufacturing LPS of SMEs	0.01	0.04	0.03	0.04	-	0.07	0.06	-0.01
Manufacturing LPS of LEs	0.02	0.07	0.06	0.06	-	-0.01	0.02	0.03
Other territories	0.07	-0.04	0.01	-0.07	0.01	0.01	0.02	-0.01

Source: Authors elaboration with ASIA and ISTAT data.

Econometric evidence of dynamic coupling

Model estimates in Table 3 further support a trend towards a dynamic coupling between KIBS and manufacturing, and therefore a propensity towards territorial servitisation. In fact, the time-interacted terms reveal significant positive coefficients for all manufacturing LPS LQs when regressed on the aggregate KIBS LQs. Specifically, dynamic coupling is observed between P-KIBS and all manufacturing LPSs, but only in low-tech manufacturing LPSs for T-KIBS. More in detail, this suggests that firms in low-tech manufacturing sectors are increasingly relying on external, technological specialised service competences, such as those offered by T-KIBS to maintain or build their competitiveness. One can infer that T-KIBS compensate for the lack of internal competences that such firms have, and our evidence suggests that their greater demand is translating into more T-KIBS locating in these manufacturing areas. In other words, we find that the LPSs are able to create or attract within their local areas the necessary KIBS competence. This points to a tension towards territorial servitisation.

Conclusion

The article makes a novel contribution to the literature on territorial servitisation and addresses a gap in the existing literature by looking at the territorial mapping and dynamics between manufacturing industries and KIBS in different types of LPSs, in the case of Italy. To this end, the article links the theoretical debate on territorial servitisation with a stream of the literature that looks at servitisation as a process, specifically involving digital industries. Drawing on empirical static and dynamic analyses, the spatial distribution of KIBS (P-KIBS and T-KIBS) and manufacturing activities is mapped out across different types of LPSs in Italy and tracked over time looking at the period 2012–18.

The article’s main findings are threefold. First, the presence of KIBS in manufacturing LPSs and in contiguous ones suggests that the spatial scale of territorial servitisation needs to be rethought, leading to a novel conceptualisation of spatial contiguity in relation to territorial servitisation. In fact, our findings suggest that a more elastic and flexible spatial radius between KIBS and manufacturing firms should be considered. We find permeable borders between urban and sub-urban areas, allowing manufacturing firms to be in the catchment area of KIBS. In part, this can be explained by the density and spatial continuity of manufacturing activities across some Italian regions in a continuum between cites (KIBS) and suburbs (manufacturing industries). Second, we find a territorial decoupling with KIBS located in urban areas and manufacturing activities mostly located outside them. Finally, despite the territorial decoupling, the dynamic analysis finds evidence of a process of territorial servitisation with KIBS and manufacturing spatially converging overtime, especially for low-tech manufacturing activities. The fast growth of T-KIBS and P-KIBS in manufacturing LPSs might be read as evidence of an acceleration in the process of territorial servitisation and a dynamic coupling process between manufacturing and KIBS, including T-KIBS. This suggests that the presence of KIBS is intensifying close to a growing demand for knowledge-intensive competences especially from such industries. In the Italian economy, low-tech manufacturing industries include globally competitive high-design, highly value-added and top-end niche products. On reflection, therefore, this result should not surprise: the competitive advantage Italian industries have in such segments requires access to expert competences that only KIBS can offer because of being substantially distant from their specialisation.

The findings of the article suggest two important policy implications. One regards the nature of industrial policies that could support the process of servitisation manufacturing that firms must undergo to maintain or build their competitive advantage.

A recoupling process between KIBS and manufacturing industries is found to be endogenous to the extent that LPSs seem capable of developing KIBS internally to the system or to access contiguous ones. Given that servitisation processes are found to differ depending on the makeup of the local economies (industrial specialisation, dominant size of firms), bottom-up and place-based industrial policies are better suited to activate instruments that targets specifically what KIBS local manufacturing forms need to access and how. While internalisation processes of servitisation seem to involve larger firms, small- and medium-sized firms are more reliant on KIBS, resulting in evidence of territorial servitisation. In Italy, industrial development policy is already place-based given the EU-wide framework of regional policy it works with, allowing for a more dispersed and capillary intervention across the many types of LPSs that a country can have (European Commission, 2024). In addition to supporting the competitiveness of manufacturing enterprises, recoupling can help avoid or alleviate existing disparities between different territories in terms of standards of living, and thus can foster fairer and more sustainable development processes. However, a more policy centralised country, such as the UK, would rely more on one-fit-all policy tools that impact idiosyncratically across localities with uneven results, for example, favouring some industries more than others.

The other relevant policy implication is that 4.0 technologies are impacting on both high-tech and low-tech manufacturing industries alike, and therefore, processes of territorial servitisation regarding digital technologies (T-KIBS) are not exclusive to high-tech manufacturing industries (as Industry 4.0 literature often argues). Rather their competences are critical for the business model transformation of low-tech manufacturing activities, especially if positioned at top-end market niches. Here competition is mostly price-inelastic and reliant on customisation, design and quality (Vendrell-Herrero et al., 2024). The ‘Made in Italy’ industries are the most dynamic component of globally competitive Italian exports, and they have therefore a greater incentive to understand and adopt KIBS expertise that ensure their advantaged in global markets (Giglioli and Giordano, 2023). This implies that a policy focus should prioritise raising awareness of the benefits of T-KIBS and facilitating access and linkages with them, especially when this means operating across localities (albeit in the same region). Connecting urban T-KIBS and sub-urban low-tech manufacturing firms should be prioritised (knowledge bridging) by local stakeholders to ensure the longer-term competitiveness of both. This again can be better achieved by a place-based industrial policy that has a regional focus and is able to identify such key connections and activate them. In other countries, like the UK, other industries show greater openness to servitisation, including aerospace, automotive or energy (oil and gas and wind). These industries are present unevenly across regions and distant from the largest concentration of KIBS, which is London; therefore, a centralised policy approach to accelerate the servitisation process would fall short of placing sufficient emphasis on a more capillary presence of KIBS nearer to industries and more targeted to the needs of the specific industry.

While addressing an important gap in the current debate, the findings of the article leave some questions open for further research. For instance, an even more fine-grained empirical analysis would shed light on what specific T-KIBS activities are more relevant for which manufacturing industry, or expanding the analysis beyond manufacturing, the current debate on green transition is, for example, bringing agriculture back to the fore of our economies, and it would be insightful to better understand the impact of P-KIBS and T-KIBS on agricultural industries.

Footnotes

Appendix 1

Table 8.

Aggregation of manufacturing based on NACE Rev.2.

High-/medium-high-tech manufacturing	Low-/medium-low-tech manufacturing
20 – Manufacture of chemicals and chemical products 21 – Manufacture of basic pharmaceutical products and pharmaceutical preparations 26 – Manufacture of computer, electronic and optical products 27 – Manufacture of electrical equipment 28 – Manufacture of machinery and equipment n.e.c. 29 – Manufacture of motor vehicles trailers and semi-trailers 30 – Manufacture of other transport equipment	10 – Manufacture of food products 11 – Beverages 12 – Tobacco products 13 – Textile 14 – Wearing apparel 15 – Leather and related products 16 – Wood and of products of wood 17 – Paper and paper products 18 – Printing and reproduction of recorded media 19 – Manufacture of coke and refined petroleum products 22 – Manufacture of rubber and plastic products 23 – Manufacture of other non-metallic mineral products 24 – Manufacture of basic metals 25 – Manufacture of fabricated metals products, excepts machinery and equipment 31 – Manufacture of furniture 32 – Other manufacturing 33 – Repair and installation of machinery and equipment

High-/medium-high-tech manufacturing

Low-/medium-low-tech manufacturing

20 – Manufacture of chemicals and chemical products
21 – Manufacture of basic pharmaceutical products and pharmaceutical preparations
26 – Manufacture of computer, electronic and optical products
27 – Manufacture of electrical equipment
28 – Manufacture of machinery and equipment n.e.c.
29 – Manufacture of motor vehicles trailers and semi-trailers
30 – Manufacture of other transport equipment

10 – Manufacture of food products
11 – Beverages
12 – Tobacco products
13 – Textile
14 – Wearing apparel
15 – Leather and related products
16 – Wood and of products of wood
17 – Paper and paper products
18 – Printing and reproduction of recorded media
19 – Manufacture of coke and refined petroleum products
22 – Manufacture of rubber and plastic products
23 – Manufacture of other non-metallic mineral products
24 – Manufacture of basic metals
25 – Manufacture of fabricated metals products, excepts machinery and equipment
31 – Manufacture of furniture
32 – Other manufacturing
33 – Repair and installation of machinery and equipment

Source: Eurostat, 2020 High-tech industry and Knowledge-intensive services, Annex 3. https://ec.europa.eu/eurostat/cache/metadata/Annexes/htec_esms_an3.pdf.

Appendix 2

Figure 6 presents the scatterplots between LQ KIBS and LQ MAN (high- and low-tech combined) for every year of the panel by type of LPS.

While for LUA and other LPSs, the relationship between LQ KIBS and LQ MAN seems not to follow a clear pattern, scatterplots for LE and SME point to a seemingly linear relationship between the two indicators.

Disclaimer

The authors are solely responsible for the content of the article. The views expressed are purely those of the authors and may not in any circumstances be regarded as stating an official position of the European Commission.

Data availability statement

The data that support the findings of this study are openly available on the I. Stat platform of the Italian National Institute of Statistics at https://siqual.istat.it/SIQual/visualizza.do?id=8889016&refresh=true&language=UK, and at .

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Edoardo Baldoni

Notes

References

Albors-Garrigos

Hervás-Oliver

(2012) Radical innovation and technology diffusion in traditional clusters: how high-tech industries reinvented a traditional cluster. In: TG

Bas

(ed.) Comparing High Technology Firms in Developed and Developing Countries: Cluster Growth Initiatives. Palmdale, PA: IGI Global, pp. 99–110.

Anselin

(2019) Global spatial autocorrelation (3). Bivariate, differential and EB rate Moran scatter plot. In: GeoDa–An Introduction to Spatial Data Science. Available at: https://geodacenter.github.io/workbook/5b_global_adv/lab5b.html (accessed 30 April 2024).

Anselin

(2020) Local spatial autocorrelation (3). Multivariate local spatial autocorrelation. In: GeoDa–An Introduction to Spatial Data Science. Available at: https://geodacenter.github.io/workbook/6c_local_multi/lab6c.html (accessed 30 April 2024).

Araya

Horváth

Leiva

(2020) The role of county competitiveness and manufacturing activity on the development of business service sectors: a precursor to territorial servitization. Investigaciones Regionales–Journal of Regional Research 2020/3(48): 19–35.

Aslesen

Isaksen

(2007) New perspectives on knowledge-intensive services and innovation. Geografiska Annaler: Series B, Human Geography 89(Suppl. 1): 45–58.

Baines

Bigdelia

Sousab

Schroeder

(2020) Framing the servitization transformation process: a model to understand and facilitate the servitization journey. International Journal of Production Economics 221(6): 1–43.

Baines

Lightfoot

Benedettini

(2009) The servitization of manufacturing: a review of literature and reflection on future challenges. Journal of Manufacturing Technology Management 20(5): 547–567.

Baines

Lightfoot

Evans

(2007) State-of-the-art in product-service systems. Proceedings IMechE Part B: Journal of Engineering Manufacturing 221(10): 1543–1552.

Baines

Ziaee Bigdeli

Bustinza

Shi

Baldwin

Ridgway

(2017) Servitization: revisiting the state-of-the-art and research priorities. International Journal of Operations & Production Management 37(2): 256–278.

10.

Baldoni

(2023) AGPRIS: Agricultural Productivity in Space (version 2.0). Available at: https://CRAN.R-project.org/package=AGPRIS

11.

Baldoni

De Propris

Storai

(2022) Territorial servitisation in Italian industrial districts: the role of technological and professional KIBS. Competitiveness Review: An International Business Journal 32(5): 743–759.

12.

Baldwin

(2016) The Great Convergence: Information Technology and the New Globalization. Cambridge, MA: Harvard University Press.

13.

Becattini

Bellandi

De Propris

(2009) A Handbook of Industrial Districts. Cheltenham: Edward Elgar.

14.

Bellandi

De Propris

Vecciolini

(2021) Effects of learning, unlearning and forgetting on path development: the case of the Macerata-Fermo footwear industrial districts. European Planning Studies 29(2): 259–276.

15.

Billé

Tomelleri

Ravazzolo

(2023) Forecasting regional GDPs: a comparison with spatial dynamic panel data models. Spatial Economic Analysis 18(4): 530–551.

16.

Bouayad-Agha

Turpin

Védrine

(2013) Fostering the development of European regions: a spatial dynamic panel data analysis of the impact of cohesion policy. Regional Studies 47(9): 1573–1593.

17.

Brittes Benitez

Fabián Ayala

Frank

(2020) Industry 4.0 innovation ecosystems: an evolutionary perspective on value creation. International Journal of Production Economics 228(C): 107735.

18.

Brunow

Hammer

McCann

(2020) The impact of KIBS’ location on their innovation behavior. Regional Studies 54(9): 1289–1303.

19.

Bustinza

Opazo-Basaez

Tarba

(2022) Exploring the interplay between Smart Manufacturing and KIBS firms in configuring product-service innovation performance. Technovation 118(C): 102258.

20.

Capello

Lenzi

(2021) The Regional Economics of Technological Transformations. London: Routledge.

21.

Capello

Caragliu

Lenzi

(2012) Is innovation in cities a matter of knowledge-intensive services? An empirical investigation. Innovation: The European Journal of Social Science Research 25(2): 151–174.

22.

Chapain

Cooke

De Propris

MacNeill

Matteos-Garcia

(2010) Creative clusters and innovation (final report). Available at: nesta.org.uk (accessed 10 April 2023).

23.

Cooke

(2002) Knowledge Economies: Clusters, Learning and Cooperative Advantage. London: Routledge.

24.

Cooke

De Laurentis

Tödtling

Trippl

(2007) Regional Knowledge Economies. Cheltenham: Edward Elgar Publishing.

25.

De Propris

(2005) Mapping local production systems in the UK: methodology and application. Regional Studies 39(2): 197–211.

26.

De Propris

Bailey

(2020) Disruptive industry 4.0+: key concepts. In: De Propris

Bailey

(eds) Industry 4.0 and Regional Transformations. London; New York: Taylor and Francis, pp. 1–23.

27.

De Propris

Bellandi

(2021) Regions beyond Industry 4.0. Regional Studies 55(10/11): 1609–1616.

28.

De Propris

Storai

(2019) Servitizing industrial regions. Regional Studies 53(3): 388–397.

29.

Dimache

Roche

(2013) A decision methodology to support servitization of manufacturing. International Journal of Operations & Production Management 33(11/12): 1435–1457.

30.

Elhorst

(2010) Dynamic panels with endogenous interaction effects when T is small. Regional Science and Urban Economics 40(5): 272–282.

31.

Elhorst

Zandberg

De Haan

(2013) The impact of interaction effects among neighbouring countries on financial liberalization and reform: a dynamic spatial panel data approach. Spatial Economic Analysis 8(3): 293–313.

32.

European Commission (2012) Knowledge Intensive (Business) Services in Europe. Brussels: European Commission.

33.

European Commission (2024) Italy 2024 digital decade country report. Available at: https://digital-strategy.ec.europa.eu/en/factpages/italy-2024-digital-decade-country-report

34.

Fingleton

(2024) A spatial econometric analysis of productivity variations across US cities. International Regional Science Review 47(4): 475–508.

35.

Fischer

(2021) The bias of avoiding spatial dynamic panel models: a tale of two research teams. Region 8(1): 153–180.

36.

Florida

(2002) The Rise of the Creative Class: And How It’s Transforming Work, Leisure, Community and Everyday Life. New York: Basic Books.

37.

Gereffi

Humphrey

Sturgeon

(2005) The governance of global value chains. Review of International Political Economy 12(1): 78–104.

38.

Giglioli

Giordano

(2023) Exploring the recent resilience of Italy’s goods exports: competitiveness, energy intensity and supply bottlenecks. Bank of Italy Occasional paper no. 785, 25 July. Rome: Banca d’Italia.

39.

Glaeser

(1998) Are cities dying? Journal of Economic Perspectives 12(2): 139–160.

40.

Glaeser

(2000) The new economics of urban and regional growth. In: Clark

Gertler

Feldman

(eds) The Oxford Handbook of Economic Geography. Oxford: Oxford University Press, pp. 83–98.

41.

Halleck Vega

Elhorst

(2013) Modelling regional labour market dynamics in space and time. Papers in Regional Science 93(4): 819–841.

42.

Hauknes

Knell

(2009) Embodied knowledge and sectoral linkages: an input–output approach to the interaction of high- and low-tech industries. Research Policy 38(3): 459–469.

43.

Houghton

Sheehan

(2000) A Primer on the Knowledge Economy. Melbourne, VIC, Australia: Victoria University.

44.

ISTAT (2014) I sistemi locali del lavoro 2011. Available at: https://www.istat.it/it/archivio/142676

45.

ISTAT (2015) I distretti industriali 2011. Available at: https://www.istat.it/it/archivio/150320

46.

ISTAT (2019) Aggiornamento dei Sistemi Locali del Lavoro 2011 a seguito di fusioni di comuni. Available at: https://www.istat.it/wp-content/uploads/2024/03/Modifiche_SLL_2011_REV_04_06_2019.pdf

47.

Lafuente

Vaillant

Vendrell-Herrero

(2017) Territorial servitization: exploring the virtuous circle connecting knowledge-intensive services and new manufacturing businesses. International Journal of Production Economics 192(C): 19–28.

48.

Lafuente

Vaillant

Vendrell-Herrero

(2019) Territorial servitization and the manufacturing renaissance in knowledge-based economies. Regional Studies 53(3): 313–319.

49.

Lafuente

Vaillant

Vendrell-Herrero

(2023) Product-service innovation systems: opening-up servitization-based innovation to manufacturing industry. Technovation 120: 102665.

50.

Lombardi

Santini

Vecciolini

(2022) Drivers of territorial servitization: an empirical analysis of manufacturing productivity in local value chains. International Journal of Production Economics 253(C): 108607.

51.

Marino

Trapasso

(2020) Servitization and territorial self-reinforcing mechanisms: a new approach to regional competitiveness. Investigaciones Regionales–Journal of Regional Research 2020/3(48): 85–93.

52.

Miles

Snow

(1984) Fit, failure and the hall of fame. California Management Review 26(3): 10–28.

53.

OECD (2004) Innovation in the Knowledge Economy: Implications for Education and Learning. Paris: OECD Publishing.

54.

Opazo-Basáez

Narvaiza Cantín

Campos

J-A

(2020) Does distance really matter? Assessing the impact of KIBS proximity on firms’ servitization capacity. Investigaciones Regionales–Journal of Regional Research 2020/3(48): 51–68.

55.

Opazo-Basáez

Vendrell-Herrero

Bustinza

Raddats

(2024) Guest editorial: digital service innovation: ontology, context and theory. Journal of Service Management 35(2): 129–140.

56.

Rios

(2017) What drives unemployment disparities in European regions? A dynamic spatial panel approach. Regional Studies 51(11): 1599–1611.

57.

Rooney

Hearn

Ninan

(eds) (2005) Handbook on the Knowledge Economy. Cheltenham: Edward Elgar Publishing.

58.

Sforzi

Boix

(2019) Territorial servitization in Marshallian industrial districts: the industrial district as a place-based form of servitization. Regional Studies 55(10–11): 398–409.

59.

Shih

(1996) Me-too Is Not My Style: Challenge Difficulties, Break through Bottlenecks, Create Values. Taipei: The Acer Foundation.

60.

Vaillant

Lafuente

Horvát

Vendrell-Herrero

(2021) Regions on course for the Fourth Industrial Revolution: the role of a strong indigenous T-KIBS sector. Regional Studies 55(10–11): 1816–1828.

61.

Vandermerwe

Rada

(1988) Servitization of business: adding value by adding services. European Management Journal 6(4): 314–324.

62.

Vendrell-Herrero

Wilson

(2017) Servitization for territorial competitiveness: taxonomy and research agenda. Competitiveness Review: An International Business Journal 27(1): 2–11.

63.

Vendrell-Herrero

Para-González

Mascaraque-Ramírez

Freixanet

(2024) The order of the factors matters: how digital transformation and servitization integrate more efficiently. International Journal of Production Economics 271(C): 109228.

64.

Vrontis

Tardivo

Bresciani

Viassone

(2018) The competitiveness of the Italian manufacturing industry: an attempt of measurement. Journal of the Knowledge Economy 9(4): 1087–1103.

65.

Wanzenböck

Piribauer

(2016) R&D networks and regional knowledge production in Europe: evidence from a space-time model. Papers in Regional Science 97(51): S1–S24.

66.

Xing

Liu

Davies

(2023) Servitization innovation: a systematic review, integrative framework, and future research directions. Technovation 122(C): 102641.

67.

de Jong

Lee

(2008) Quasi-maximum likelihood estimators for spatial dynamic panel data with fixed effects when both n and T are large. Journal of Econometrics 146(1): 118–134.