J Evol Econ (2009) 19:173–196 DOI 10.1007/s00191-008-0128-2 REGULAR ARTICLE
Modes of innovation in knowledge-intensive business services evidence from Lombardy Nicoletta Corrocher · Lucia Cusmano · Andrea Morrison
Published online: 25 September 2008 © Springer-Verlag 2008
Abstract The present paper investigates the sectoral variety and common patterns across different typologies of knowledge-intensive business services (KIBS). We examine this issue by considering the case of Lombardy, a highly developed manufacturing area the industrial activities of which are experiencing a pervasive transformation towards higher knowledge content, demanding stronger and more pervasive support by advanced services. Drawing on an original survey-based firm-level dataset, we analyze innovation patterns across KIBS, evaluating the explanatory power of traditional classifications of the service sector, as well as the heterogeneity driven by firm and market specific characteristics. Our findings highlight four profiles of KIBS: interactive innovation mode, product innovation mode, conservative innovation mode and techno-organizational innovation mode. When examining in more depth the variables that are associated with cluster membership, we find that firm strategy is the most significant determinant, with size, customer location, and training also playing a role in defining cluster specificities.
N. Corrocher (B) · L. Cusmano · A. Morrison CESPRI, Bocconi University, Via Sarfatti 25-20236, Milan, Italy e-mail:
[email protected] N. Corrocher Department of Economics, NFH, University of Tromso, 9037 Tromso, Norway L. Cusmano Insubria University, Via Monte Generoso 71, 21100 Varese, Italy A. Morrison Department of Economic Geography, Utrecht University, 3508 TC, Utrecht, The Netherlands
174
N. Corrocher et al.
Keywords Innovation modes in services · KIBS · Italy JEL Classification O31 · O50 · L84
1 Introduction: knowledge-intensive business services in the knowledge economy The attention paid to the innovative activities of service sectors has significantly increased over the last decade. Since the seminal work by Miles et al. (1995) an increasing number of contributions have questioned the traditional view of service firms as incapable of producing innovations. Scholars and practitioners have acknowledged that, far from being innovative laggards or just intensive users of technologies and novelties generated by manufacturing, services are becoming an ever more important locus for innovative activity (Howells 2000; Tether and Metcalfe 2004). This is particularly true for ‘knowledge-intensive business services’ (KIBS), such as those involved in consultancy, market research, design, engineering and technical services, the role of which in the dynamics of modern ‘knowledge economies’ extends well beyond their actual direct employment relevance (Muller and Zenker 2001; Gallouj 2002; Miles 2005). As Doloreux and Muller (2007) describe it, the perception of KIBS has evolved from an early characterization as providers or transferers of specific information for their clients (Wood et al. 1993) to the recent identification as key nodes of knowledge-related networks, which can trigger and strengthen processes of knowledge conversion in client firms (den Hertog 2000). The role of KIBS appears to be particularly significant in advanced regions, where manufacturing competitiveness increasingly depends on knowledge content, provided by highly specialized suppliers. In this respect, the innovation literature has been emphasizing the novelty brought about in advanced economies by these ‘bridges of innovation’, which interact with the manufacturing sector as knowledge purchasers, providers and partners (Czarnitzki and Spielkamp 2000). Tether and Hipp (2002) suggest that, indeed, the recent tertiary evolution and outsourcing trends, concerning both routine activities and innovation tasks, imply a redistribution of knowledge in favour of KIBS and away from traditional producers and service providers. In the empirical literature, attention has been mainly directed at detecting service-specific modes of innovation, as opposed to manufacturing-specific ones. When focusing exclusively on the tertiary sector, empirical studies have investigated the distinct role and different innovation practices of traditional vs. knowledge-intensive service activities (Evangelista and Savona 2003; Tether et al. 2001; Hollenstein 2003; Tether 2005; Freel 2006). It is KIBS in particular which are proposed as the most evident counterexample to the generic classification of tertiary activities into Pavitt’s taxonomy of ‘supplier dominated’ sectors (Tether et al. 2001; Toivonen 2004; Camacho and Rodriguez 2005; Freel 2006). Although these contributions have had a major role in enlightening the different innovative approaches in manufacturing and
Modes of innovation in knowledge-intensive business services. . .
175
services at large, and emphasizing the peculiarity of KIBS among tertiary activities, such a perspective neglects the remarkable heterogeneity within this same largely defined KIBS category. In other words, much of the attention has been paid to differentiating them from traditional services. Therefore “a serious challenge remains to unpack these different types of services and to explore their different approaches to innovation” (Tether 2005: 180). The present paper intends to explore the KIBS’ ‘black box’, investigating sectoral variety and common patterns across different typologies of this generally defined service category. Some recent empirical work (e.g., Freel 2006) provides detailed evidence on innovation patterns at the firm level, portraying differences between technical KIBS (t-KIBS: IT related services, engineering, R&D consulting, etc.) and professional services (p-KIBS: business and management services, legal and accounting activities, market research, etc.). The classification into these main sub-sectors is gradually taking root in the literature (Miles et al. 1995; Freel 2006; Doloreux and Muller 2007). This is certainly useful, but, when taken as an a-priori distinguishing factor, it can also prevent detection of other relevant differences within the same category or common patterns across differently classified types of activities. We believe that investigating patterns in the data, beyond a-priori sectoral distinctions, is all the more relevant as the empirical research on the issue is in its early stage. We examine the issue by considering the case of Lombardy, a highly developed manufacturing area the industrial activities of which are experiencing a pervasive transformation towards higher knowledge content, demanding stronger and more pervasive support by advanced services. Drawing on an original survey-based firm-level dataset, we therefore explore innovation patterns across KIBS, evaluating the explanatory power of traditional classifications, as well as heterogeneity driven by firm and market specific characteristics. The paper is organized as follows. Section 2 briefly discusses factors for competitiveness and patterns of innovation in KIBS, as emerging from the literature, and the state of the art of the empirical research. Section 3 describes the KIBS sector in Lombardy, outlines the methodology of investigation and provides some descriptive evidence on the sample of analysis. Section 4 is the core of the paper. First, it presents the results of a factor analysis addressing innovation patterns in KIBS. Second, it illustrates the results of a cluster analysis carried out on the previously identified innovative factors. Differences across clusters are discussed and assessed in relation to a set of firm-level and market characteristics, by means of a multinomial logistic regression. Section 5 concludes, providing interpretations of the main results.
2 Competition and innovation in KIBS: the main issues Over the last decade, the economic and business literature has been largely discussing competitive strategies and innovation patterns in KIBS, both from a theoretical perspective and, to a lesser extent, from an empirical point of view. Commentators have been mainly focusing on what makes KIBS’ competitive
176
N. Corrocher et al.
approach and innovation routines different from other more traditional service activities and the manufacturing sector at large. User participation in the process of production and delivery, often overlapping with consumption itself, is generally the most debated and distinguishing characteristic of services and appears all the more relevant in the case of knowledge intensive services (Barras 1990; Gadrey and Gallouj 1998; Sundbo and Gallouj 2000). Tether et al. (2001) notice that the extent to which the user is involved in the provision of the service varies broadly, from the service being carried out on behalf of the user by the provider, to the service being carried out by the user with the facilities or the equipment of the service provider. In this respect, there exists a tension between the pressure to reduce the production costs of services, which leads firms to look for increasing standardization, and the need to meet specific user requirements, which, on the contrary, force firms to seek a high degree of customization in their products. Simultaneity of production and consumption and the intangible nature of the service make long distance trade more difficult than for other goods and give a local flavour to competition, even when considering the more sophisticated services (de Jong et al. 2003). Koschatzky and Zenker (1999) underline the idea that KIBS’ intensive interaction with clients and their knowledge environment are not spatially neutral. Proximity matters, as complex cognitive processes need not only large flows of codified scientific technical information, but also tacit knowledge for using and interfacing that information. The relevance of user-producer interaction is also associated with the innovative dimension. The literature often stresses the fact that KIBS are involved in interactive learning processes both with their customers and with other organizations within the local innovation system (Strambach 1998; den Hertog 2000; Thomi and Böhn 2003). KIBS provide non-material knowledge-intensive services to client firms or institutions, possibly generating new knowledge as a result of the interaction with the customers. Gallouj and Weinstein (1997) and Kuusisto and Viljamaa (2004) explicitly refer to ‘co-production’, since service providers and customers bring different sets of capabilities and competencies. Following this line of reasoning, Gadrey and Gallouj (1998) emphasize that, in professional services, the provider-customer interface represents both a locus and a source of innovation, particularly for ad hoc innovations. According to Muller and Zenker (2001), the development of KIBS’ knowledge base is intimately related to the activity they perform for their clients, and this mutual contribution defines a knowledge-base loop. Similarly, Tether and Metcalfe (2004) argue that cooperation with customers and suppliers represents for services the main source of knowledge and technology, a ‘soft’ source which defines their innovation strategy more than traditional ‘hard’ ones such as R&D activities. Following the dichotomy proposed by Jensen et al. (2007), the mode of learning and innovation in services can be described as ‘doing, using and interacting’ (the so-called DUI model), relying on experience-based know-how, rather than upon the production and use of codified scientific and technological knowledge (the so-called STI mode). Recent studies (Coombs and Miles 2000;
Modes of innovation in knowledge-intensive business services. . .
177
Tether 2005; Freel 2006) provide strong support for this characterization of innovative patterns in relation with KIBS, highlighting interactive processes and the intangible nature of input and output, which, however, also makes it hard to measure innovation itself (Gallouj and Weinstein 1997). Cooperation with other service firms is another potential source of innovation for KIBS (Bryson and Monnoyer 2004), which, however, tends to be constrained by appropriability concerns, that is, by the weakness of IPR protection (Freel 2006). Although it is largely acknowledged that the interaction with users and other firms is a relevant source for innovation in KIBS, several scholars have highlighted the accumulation of competencies from a variety of different sources and the existence of multiple patterns of innovation that are not highly interactional (Sundbo and Gallouj 2000; den Hertog 2000).1 One such non-interactive source of knowledge is given by the adoption of technology. Relation with the ICT is a key feature of KIBS which has been attracting a great deal of attention in the literature. In the most dynamic service industries, investments in ICT are larger than in the manufacturing sector and the available statistics concerning ICT investments show that service sectors account for the biggest - and growing - share of the total expenditures in ICT in the economy. The fact that services are highly related with the ICT has another important implication, which concerns the modes and timing of production and delivery. In principle, the emergence of new technologies makes the ordering, reservation and delivery of some types of services possible and easier. This process can introduce some distance between service development and utilization. This can increase the geographical reach of KIBS and, accordingly, the perception of an increasing international pressure on local firms. This phenomenon is ever more evident in service areas such as consulting, banking and logistics, where new information and communication technologies have given firms a global scope of actions. In addition, a sharper separation between the development and the utilization of the service can increase specialization, with some KIBS focusing on either one or the other stage. Besides, this outcome is likely to reinforce the distinct patterns of innovation already observed in different kinds of KIBS (Miles et al. 1995), with professional KIBS more keen to adopt new technologies, while technical KIBS more focused on moulding them. A third feature of services, which affects the development of innovative activity, is their labour-intensive nature (Sirilli and Evangelista 1998; Licht and Moch 1999). It is especially in KIBS that highly qualified human capital represents a key strategic asset.2 Investments in human resources appear to increase, rather than decrease, with the introduction of ICT, following the need for firms to improve their capacity to harness technology. Indeed, as the complexity of technology increases, so does the need for intensive learning,
1 According
to Tether (2003), the emphasis on the ‘interaction’ side is also to be understood in relation to the aim of reversing the traditional view of supplier-dominated services. 2 Recent evidence supports this claim. For example, Freel (2006) shows that KIBS tend to employ high skill workers, and technical KIBS in particular hire high profile technicians.
178
N. Corrocher et al.
which helps transform largely available information into valuable idiosyncratic knowledge. Despite the increasing knowledge codification, deriving from the introduction of ICT in the work routines, the tacit component of knowledge still plays a relevant role in the elaboration of the information and in its transformation into value-added content for new services. Accordingly, human resources are increasingly demanded to take part in the creation of services’ knowledge content and, in particular, in the customization of the service provision. In relation to this, promising fields for innovation also exist in the non-technological KIBS (Toivonen 2004). The transition to knowledge-based, service-oriented economies therefore raises the importance of human capital, and the skill shortage in such sectors has already emerged as one of the major problems for developed countries (Bresnahan et al. 2002). As far as the direction of innovative efforts is concerned, conceptualization and empirical analysis have mainly focused on what differentiates services from manufacturing. Tether (2005) contends that innovation in services, as opposed to manufacturing, is more likely to entail and be oriented towards organizational change, rather than product/process innovation. Even though experts skills are very important for competitive service provision and highly qualified staff tends to be associated with higher levels of innovativeness (Freel 2006), service innovation appears to be strongly related with organizational learning and knowledge. Larsen (2001) defines a ‘distributed knowledge system view’, arguing that, rather than the sum of internally available resources, it is the way the employees interact socially with internal and external colleagues (taking part into ‘communities’) which determines KIBS’ knowledge base. As Leiponen (2001) points out, it is especially in KIBS that the innovative efforts target the organizational level, aiming to a significant degree at standardizing services and underlying procedures. The evidence on the relationship between KIBS’ innovation and standardization is, however, not conclusive. Hipp et al. (2000), for instance, find that the pattern of innovation is inversely related to the degree of standardization of services: firms providing standardized services tend to be less innovative than firms offering customized services. Tether and Hipp (2002) argue that, indeed, a high level of investment in ICT and service customisation is a distinct feature of technical KIBS. Heterogeneity of innovative approaches in services has been increasingly acknowledged and empirically explored (Sundbo and Gallouj 2000; den Hertog 2000). In recent empirical contributions that investigate such a variety of tertiary innovation modes, however, KIBS are generally intended as a homogeneous category, which is confronted with other service activities. The works by Tether (2003) and Freel (2006) provide important steps in the direction of exploring differences across KIBS. According to these authors, innovation strategies are much related to competitive circumstances, which represent a more relevant differentiating factor than other more objectively measurable variables, such as size or industry classification. Yet, empirical studies in this direction perform comparisons based on nomenclature classification (ISIC or NACE) or on the a-priori distinction between professional and technical services (Toivonen 2004), as firstly proposed by Miles et al. (1995).
Modes of innovation in knowledge-intensive business services. . .
179
The work by Hollenstein (2003) represents an attempt to investigate modes of innovation in services, irrespective of a priori nomenclature. Analyzing the Swiss service sector with firm-level data, Hollenstein (2003) proposes a clusterization of firms which cuts across traditionally defined categories of tertiary activities. Five different clusters of service firms are identified, ranging from high tech firms with a science-based profile and full integration into knowledge networks, to low profile innovators with almost no external link. Hollenstein’s work provides methodological novelty, but does not concern KIBS firms in particular. The KIBS’ ‘black box’ has been opened following nomenclature distinctions, such as the increasingly popular technical vs. professional dichotomy (Doloreux and Muller 2007). On the base of firm-level data, we first explore modes of innovation in KIBS along such a dichotomy, but we then cut across a priori sectoral distinctions, testing the relevance of firm- and market- specific characteristics in explaining differentiated attitudes towards innovation.
3 The KIBS sector in Lombardy: data and descriptive evidence Over the last fifteen years, Lombardy has undergone an important process of structural change, from an industry-based to a service-based economy. Notwithstanding its traditional orientation towards industrial activities (38% of total employees in 2001), the region has experienced an increase in value added and employment in the service sector. Between 1991 and 2001, these grew by 2.4% and 13.2%, respectively, as compared to an Italian average of 2.3% and 9.2%. This process has been particularly pronounced in KIBS and financial services. In particular, the growth of the KIBS sector has been well above the tertiary industry average, with a growth rate of 121.3% for local units and 94.5% for employees. In terms of weight in the regional economy, the KIBS sector’s relevance grew both in terms of local units (from 12% to 23.1% of the regional total) and in terms of employees (from 7.8% to 14.4%). The empirical investigation mirrors the current structure of this evolving sector in Lombardy. It relies upon a stratified sample of 441 KIBS, which is representative of the regional universe.3 The stratification is based on two variables: geographical location and sector. Four main geographical areas have been considered through the aggregation of the Lombardy provinces (Milan; Varese, Como, Lecco and Sondrio; Pavia, Lodi, Cremona and Mantova; Bergamo and Brescia). However, more than 50% of firms operate in the Milan province, which is consistent with the general evidence of KIBS concentration in large metropolitan areas (Strambach 1998; Simmie and Strambach 2005). Nevertheless, the Northern and Eastern part of the region are also significantly
3 Data
collection and sampling were carried out as part of the project “Survey sulle imprese e sulla struttura economica lombarda—Settore dei servizi” sponsored by the Istituto Regionale di Ricerca della Lombardia (IRER) in 2006.
180
N. Corrocher et al.
Table 1 Survey analysis: sample composition Sectors
Descriptiona
72 721 722 723 725
Computer and related activities Hardware consultancy Software consultancy Data processing Maintenance and repair of office and computer machinery Other computer related activities Research and development Research and experimental development in natural sciences and engineering Research and experimental development in social sciences and humanities Other business activities Legal, accounting, bookkeeping, auditing activities and tax consultancy Architecture and engineering activities and other technical services Testing activities and technical analysis Advertising Other professional/business services
726 73 731
732
74 741
742
743 744 748 Total
a Classification
Number of firms
Percentage
Type
58 1 15 27 7
13.15 0.2 3.4 6.1 1.6
8 4 3
1.8 0.91 0.7
t-KIBS
1
0.2
t-KIBS
379 132
85.94 29.9
p-KIBS
156
35.4
t-KIBS
3
0.7
t-KIBS
9 79 441
2.0 17.9 100.0
p-KIBS p-KIBS
t-KIBS t-KIBS t-KIBS t-KIBS t-KIBS
adapted from Freel (2006)
represented, as they are historically characterized by high levels of industrial activity and, nowadays, ‘tertiarization’ trends. As far as sectors are concerned, the three NACE areas which are commonly associated with KIBS serve for the purpose of sample stratification (Table 1): computing services (NACE 72); research and development (NACE 73); other professional activities such as engineering offices and consulting services (NACE 74).4 The first two groups include the so-called ‘technical KIBS’, while the third, which weighs the most, is largely made of ‘professional KIBS’, such as legal, accounting, bookkeeping, auditing activities and tax consultancy (NACE 741), which account for about one third of the sample units. Following Nählinder (2002) and Freel (2006), we classify as technical KIBS (t-KIBS) ‘testing activities and technical analysis’ (NACE 743) and the highly dense subsector of architecture and engineering activities (NACE 742).
4 The second segment (R&D services) is rather sparse (four units only in the stratified sample), as a
result of the relatively low number of dedicated businesses in the regional universe. For this reason, in our descriptive analysis we consider a higher level of sector aggregation, by distinguishing between p-KIBS and t-KIBS.
Modes of innovation in knowledge-intensive business services. . .
181
The survey was carried out in spring 2006 by telephone interviews5 and aimed at capturing the most important factors for competitiveness and change at the individual entrepreneur and at the firm level. The use of survey data is particularly appropriate for investigating competitive strategies and innovation attitudes at the firm level, since, as in other types of services, innovation processes in KIBS are characterized by intangible output, strong user–supplier interaction and customization, ‘high quality labour’ intensity, and pervasive usage of ICT. In our sample, 39.5% of firms record only one employee, and 93.4% of firms fewer than ten employees. By investigating sectoral differences, we can observe that technical KIBS tend to be on average slightly smaller than professional KIBS. The size class distribution is less skewed towards micro units in the computing service sector, where nearly half of the firms fall in the four to nine employee class. The statistics about the number of employees, however, reflect the peculiar employment patterns of both technical and professional services and underestimate the actual involvement of human resources in the sector, as these activities have been always characterized by the use of a temporary labour force.6 When looking at the size evolution of the KIBS firms over the past three years, a picture of general stability emerges, especially when changes are measured in terms of employees rather than in terms of revenues. Across sectors, it is the computing service field that exhibits more pronounced employment and revenues swings. Professional KIBS are characterized by a relatively larger share of young firms: nearly half have been established within the last 10 years. Technical KIBS, by contrast, tend to be older and this is particularly evident for engineering and architecture services, which are often provided by long established businesses, although the share of very young firms (less than 5 years old) is also quite relevant (20%). In order to provide some evidence on the importance of different variables in defining competitive strategies and innovation processes of KIBS firms, we asked respondents to measure on a four-point scale the importance of a set of variables in relation to their competitive strategy (following Howells and Tether 2004) and a set of variables characterizing innovation patterns. Overall, statistically significant differences between p-KIBS and t-KIBS are detected only along a few competitive or innovative dimensions. As far as competitive strategies are concerned (Table 2), the three most important factors are quality, reputation and rapidity in service provision. This is true, on average, both for professional KIBS and for technical KIBS. When looking at less important factors, it emerges that the range of services offered appears to be more important than the introduction of new services per se, and that ‘post-sale services’ rank higher for technical KIBS than for professional KIBS.
5 The
telephone interviews were conducted by a specialized survey company with the assistance of the CATI procedure. 6 This trend has been further encouraged by the recent reforms in the Italian labour market, which introduced a large variety of flexible labour contracts.
182
N. Corrocher et al.
Technology largely matters in the provision of services, whereas ‘price’ is significantly less popular than the aforementioned quality-related factors. Distribution channels are also rarely identified as important competitive drivers. Finally, technical KIBS are more inclined than professional KIBS to consider collaboration with other firms as an important competitive variable. When considering innovation strategies, the general picture is one of KIBS that are rather reluctant at introducing novelties in the typology of services and in the modes of service delivery (Table 3), which seems to confirm the argument that process and product innovations are more common in the realm of manufacturing activities, where innovation outputs are tangible and therefore easily recorded (Tether 2005).7 Evidence in Table 3 also provides support to the view that embodied technology is crucial. The most diffused innovation investments concern the technologies employed in service provision (ICT in particular). The ‘type of services’, which represents the third most important target of innovation, and ‘modes of service production’ are the only dimensions of innovation along which a significant, though small, difference between t-KIBS and p-KIBS is recorded, with the former attributing slightly more relevance to their change. Looking at ‘softer’ sources of innovations, such as cooperation with users, along with the much claimed human resources, they do not stand out as the most peculiar aspects of Lombardy knowledge intensive services. This is somehow at odds with the arguments in the literature presented in Section 2. Nevertheless, this is not to say that softer sources of innovation are irrelevant, since a non-marginal share of KIBS (around 10%) report investing significantly in those factors. Rather, the descriptive statistics in Table 3 point at an average attitude that is not consistent with the typical characterization of KIBS, as emphasized, for instance, by Tether and Hipp (2002). We argue that this outcome might result from the averaging of rather different behaviours. It is reasonable to expect that smaller but significant sub-groups of KIBS follow patterns of innovation based on organizational change or other softer aspects of innovation. The following section specifies and details such differences, proposing a cluster analysis which cuts across traditional classifications and groups firms into homogenous categories with respect to innovation strategies.
4 Innovative patterns across KIBS: a cluster analysis In order to investigate the diversified modes of innovation within KIBS beyond the a-priori distinction between professional and technical KIBS, we
7 We
have to take into consideration the relevance of small firms in our sample. The size class skewed distribution might introduce some bias in the analysis. Nevertheless, as the KIBS sector is indeed characterized by a population of small businesses, we believe investigating alternative modes of innovation contributes to understanding the sector’s overall dynamics.
86.4 81.9 60.5 50.6 47.2 31.7 25.9 19.0 17.7 10.0 8.2
12.0 15.4 29.0 31.1 34.2 28.6 38.3 38.1 15.9 18.8 22.7
Rather important (2; %) 0.9 0.7 5.0 10.2 7.5 23.1 17.0 17.9 7.7 16.8 15.2
Not very important (3; %) 0.7 0.0 4.1 7.9 7.5 15.9 15.2 23.8 40.4 38.8 37.6
Not important at all (4; %) 1.15 1.14 1.61 1.81 1.80 2.25 2.31 2.60 3.44 3.48 3.40
P-KIBS (mean)
1.17 1.29b 1.52 1.71 1.81 2.26 2.33 2.39b 3.08b 3.15b 3.22
T-KIBS (mean)
b Statistically
significant differences between P-KIBS and T-KIBS at the 95% level.
a We asked firms to indicate on a four point scale (1 = most important, 4 = not important) to what extent different variables matter for their competitive strategies.
Quality of services Reputation Speed of service delivery Use of advanced technologies Large range of services offered Price competition Development of new services Collaboration with other firms Post-sale services Availability of distribution channels Location of distribution channels
Very important (1; %)
Table 2 Relevant factors for firms’ competitiveness (mean values by sector)a
Modes of innovation in knowledge-intensive business services. . . 183
31.7 29.9 28.1 29.9 27.7 26.1 21.8 13.2
37.4 25.4
15.6 12.9 12.9 11.6
8.4 7.0
Rather important (2; %)
25.6 21.3
26.5 25.4 17.9 29.5
14.5 21.8
Not very important (3; %)
39.5 57.6
26.5 30.6 39.7 32.2
15.2 21.8
Not important at all (4; %)
3.19 3.29
2.86 2.90 2.95 2.87
2.15 2.49
P-KIBS (mean)
3.02 3.36
2.61b 2.65b 2.85 2.81
2.07 2.38
T-KIBS (mean)
asked firms to indicate on a four point scale (1 = most important, 4 = not important) whether and to what extent their activities have changed. significant differences between P-KIBS and T-KIBS.
b Statistically
a We
Use of ICT Technologies for service production/delivery Modes of service production Types of services Human capital competences Cooperation with customers/ other firms Modes of service delivery Organizational structure
Very important (1; %)
Table 3 Types of innovation: relevance for firms’ strategiesa
184 N. Corrocher et al.
Modes of innovation in knowledge-intensive business services. . .
185
Table 4 Factors for innovation
Technologies for service production/delivery Use of ICT Human capital competences Organizational structure Modes of service production Type of services Modes of service distribution Cooperation with customers/ other firms
Technology adoption
Organizational change
Service production
External cooperation
0.86
0.17
0.22
0.11
0.85 0.30 0.03 0.30 0.24 0.12 0.21
0.17 0.75 0.73 0.16 0.47 0.04 0.47
0.20 0.08 0.34 0.81 0.64 0.45 −0.03
0.15 0.16 0.15 0.20 0.13 0.79 0.74
Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.
perform a cluster analysis on innovation variables and then compare the way in which innovative strategies vary across a set of firm- and market-specific characteristics, by means of a multinomial logistic regression. To this aim, we first reduce the number of variables relating to innovation, by way of factor analysis.8 Results are reported in Table 4. We can identify four main factors, which relate differently to the internal organization of the firm and to the relationship with other actors in the markets, such as technology providers or customers. The first factor, labelled ‘Technology adoption’, is explained by the technologies used in service production/delivery and by the use of ICT. Clearly, this group of variables characterizes firms that are at the frontier in terms of adoption and use of new technologies but that are also likely to rely upon external drivers of innovation, such as specialized suppliers of tangible technological inputs. The second factor, ‘Organizational change’, is explained by human capital competences and organizational structure, and reflects an innovative pattern which is oriented towards changing organizational variables such as the firm internal structure and personnel skills and profiles. The third factor, ‘Service production’, is explained by variables related to modes of service production and type of services and identifies firms that tend to develop innovations which are strongly market oriented. Finally, the fourth factor, ‘External cooperation’, is explained by modes of service distribution and by cooperation with customers or other firms. It captures the behaviour of firms that concentrate their innovative efforts in the ‘front office’ and in the interaction with other firms. The factor analysis provides the basic input for the cluster analysis, which aims at characterizing the variety of attitudes towards innovation across
8 We
extract 4 factors by means of a principal component analysis with Varimax rotation. The percentage of variance explained is 76.8. For the scope of the present analysis, within each component, we focus on the variables that display a factor loading greater than 0.40.
186
N. Corrocher et al.
KIBS. The purpose of the clustering exercise is to detect commonalities and differences across KIBS. This exercise then constitutes the input for a multinomial logistic regression, which aims at exploring the relevance of firmand market-specific variables in explaining different innovation modes. We perform a k-means cluster analysis based on the scores of the factor analysis. In order to determine the most appropriate number of clusters, we take into account statistical criteria, such as the relationship between within-cluster and between-cluster variance, but also the number of firms per cluster and the interpretability of the results in terms of innovation patterns. Accordingly, we extract four clusters (Table 5), whose difference in terms of factors is statistically significant. Furthermore, these clusters include a sensible number of observations and can be properly interpreted as patterns of innovation. Cluster 1 presents an innovative pattern focused on service delivery and cooperation, and so we label it as interactive innovation mode. Cluster 2 focuses on innovations related to service production (product innovation mode), Cluster 3 shows a very low propensity to innovate in all areas (conservative innovation mode), whereas Cluster 4 exhibits the most comprehensive approach to innovation, focusing on both technological innovations and organizational innovations (techno-organizational innovation mode). Tether (2005) emphasizes the need to focus on three domains in order to achieve a better conceptualization of innovation: firm output, internal organization of the firm and external organization of service provision. Interestingly, when excluding the more conservative type, our clusters of KIBS match quite well these domains. Cluster 1 is characterized by an innovation pattern focused on external cooperation and service delivery, which reflects the importance of external relations with suppliers and customers in defining service innovation patterns. Cluster 2’s innovation strategy puts emphasis on service production and is therefore close to the first domain, defined around firm outputs. Finally, firms in cluster 4 are characterized by an innovation domain revolving around techno-organizational change, which identifies the internal organization of the firms in terms of the way in which service provision is organized. The second step of our empirical analysis aims at comparing more in depth the four clusters across some key characteristics. To this aim, we estimate
Table 5 Clusters of KIBSa
Technology adoption Organizational change Service production External cooperation a Number
Cluster 1 Interactive innovation mode (86)
Cluster 2 Product innovation mode (109)
Cluster 3 Conservative innovation mode (103)
Cluster 4 Techno-organizational innovation mode (143)
−0.641 −0.335 −0.731 0.993
−0.790 0.210 0.979 −0.061
0.024 −0.675 −0.576 −1.033
0.970 0.527 0.109 0.193
of firms in parentheses.
Modes of innovation in knowledge-intensive business services. . .
187
the following multinomial logistic model, taking cluster membership as the dependent variable: K exp αi + βjk Xik k=1 Pr (Y i = j) = for j = 2, . . . , 4 J K 1 exp αh + βhk Xhk h=2
Pr (Yi = 1) =
1+
J h=2
k=1
1
exp αh +
K
for the reference category j=1 βhk Xhk
k=1
where Y is the dependent variable (in our case cluster membership), X is the vector of covariates and β is a vector of coefficients. Following the theoretical literature discussed in Section 2 and some recent empirical contributions (Hollenstein 2003; Freel 2006), we select covariates that reflect the most distinctive traits of KIBS (and services more generally). First, we consider four structural variables: sector, which is a dummy variable taking a value of 1 if the firm belongs to professional KIBS and 0 otherwise; size, which is related to the number of employees, grouped into five size classes (1; 1–3; 4–9; 10–50; >50); age, which is similarly measured in five age classes (20); geographical expansion, which is a dummy variable taking a value of 1 if the firm has offices outside the region and 0 otherwise. Second, we look at customers’ characteristics by considering two variables: customer type, which is a categorical variable, including five categories of primary customers (government/public administration, large firms, medium firms, small firms, final consumers), and customer location, an ordinal variable measured on a five-level scale, related to the distance of the main customers from the firm (same urban area, within 50 km, same region, Italy, abroad). Third, we look at variables related to the competitive environment. We consider intensity of competition, an ordered variable measured on a threepoint scale in terms of (perceived) number of competitors (1–4; 5–10; >10) and competitor location, which refers to the main competitors of the firm and is measured on a four-point scale, adopting the same distance categories used for defining customer location (same urban area, within 50 km, same region, Italy, abroad). Then, we take into account variables identifying firm competitive strategy. We start from the list of variables included in the questionnaire (see Table 2 above) and, in order to reduce their number, we perform a factor analysis. Four factors summarize the key variables for competitiveness: distribution channels, price, innovativeness and reputation (see Table 7 in the Appendix for details). The first factor, ‘Distribution channels’, captures the emphasis placed on the availability and location of distribution channels, and on post-sale services. It reflects
188
N. Corrocher et al.
the idea that, when competing in the market, firms pay attention to the interaction with customers both in terms of visibility and in terms of provision of post-sale services. The second factor, ‘Price’, is explained by variables related to price competition, speed of service delivery and range of services offered. The third factor, ‘Innovativeness’, captures the relevance given to collaboration with other firms, the use of advanced technologies and the development of new services. It characterizes firms which perceive that their competitive position depends on the technologies used and on establishing partnerships with other companies in order to develop innovations. The fourth factor, ‘Reputation’, is explained by competitive variables related to well-established brand reputation and quality of services. Factor loadings for each factor are then used as covariates to identify firm competitive strategies. Finally, we include two variables related to human resources management: new employees, which is a dummy variable taking a value of 1 if the firm hired new professional profiles over the last three years and 0 otherwise, and training, an ordinal variable which measures the participation of a firm’s personnel to training course on a three-point scale. Cluster 3, grouping firms with the most conservative approach to innovation, is taken as the base case, so that the significance of variables for other clusters should be interpreted in relation with this benchmark. Table 6 illustrates the results. First of all, it is important to notice that there are no statistically significant differences in cluster sectoral composition: professional and technical KIBS distribute evenly across the four clusters. This suggests that our clusters describe patterns of innovation that do not match with specific KIBS macrosectors, but are rather related to other firm-specific and context-specific variables. This sheds some new light on the existing discussion of innovation in KIBS, which has been lately evolving in the direction of a comparison of technical and professional KIBS. In fact, a closer investigation shows that cluster 4, despite having the highest percentage of professional KIBS, includes very different activities such as legal, accounting, bookkeeping, auditing activities and tax consultancy (sector 741), architecture and engineering activities and other technical services (sector 742), and other professional/business services (sector 748). This means that some common innovative strategies exist across technical and professional KIBS. Cluster 1—the interactive innovation cluster—records the highest percentage of technical KIBS: this might come as a surprise, if one considers that these firms have been often considered as having many similarities with high-tech manufacturing sectors (see for example Hipp et al. 2000; Freel 2006), and existing empirical evidence suggests that cooperation is generally positively associated with professional KIBS more than with technical KIBS. This apparently counter-intuitive result can be explained by the production modes and market strategies of a particular sub-sector of technical KIBS (architecture and engineering activities), which accounts for a large percentage of firms in cluster 1 (38.4%) and shares with other
Modes of innovation in knowledge-intensive business services. . .
189
Table 6 Competitive structure: differences across clustersa
Age Sector Size Distribution channels Price Innovativeness Reputation Customer location Customer type Competition Competitor location New employees Training Geographical expansion Constant Observations
Cluster 1 Interactive innovation mode
Cluster 2 Product innovation mode
Cluster 4 Techno-organizational innovation mode
0.261∗∗ (0.11) −0.123 (0.32) −0.185 (0.18) 0.659∗∗∗ (0.16) 0.0914 (0.20) −0.178 (0.19) −0.350∗ (0.18) −0.358∗∗ (0.18) 0.0815 (0.16) −0.111 (0.24) 0.244 (0.27) −0.413 (0.88) 0.178 (0.22) −1.214 (0.76) 2.897 (2.64) 441
0.159 (0.10) 0.295 (0.30) −0.00209 (0.17) 0.685∗∗∗ (0.15) 0.474∗∗∗ (0.18) 0.232 (0.17) −0.0119 (0.14) −0.0386 (0.17) 0.0488 (0.15) −0.256 (0.21) 0.0132 (0.25) 0.585 (0.98) 0.532∗∗ (0.21) 0.825 (0.96) −3.553 (3.06) 441
0.133 (0.10) 0.240 (0.30) −0.373∗∗ (0.18) 0.666∗∗∗ (0.16) 0.764∗∗∗ (0.18) 0.819∗∗∗ (0.17) −0.0793 (0.15) 0.176 (0.17) 0.189 (0.15) 0.0439 (0.22) −0.0591 (0.27) 0.466 (0.94) 0.353∗ (0.20) 1.006 (1.00) −4.130 (3.11) 441
Standard errors in parentheses. LR chi2 (42) = 164.01; Prob > chi2 = 0.0000; Log likelihood = −521.77024; Pseudo R2 = 0.1358 *** p < 0.01, ** p < 0.05, * p < 0.1. a Base case: cluster 3.
types of so-called professional activities the attitude to work in cooperation with customers and with other firms. On the other hand, KIBS related to computing activities are much less represented in the cluster. This kind of activity (i.e. computing) is indeed strongly represented in cluster 3, which is characterized by weak strategies along all the four innovation dimensions. All in all, this evidence suggests that heterogeneity in innovation strategies goes much beyond the distinction between professional and technical KIBS and that both these groups are characterized by highly idiosyncratic service activities that require further investigation. The analysis of other firm-specific variables helps characterize further the clusters. Size is significantly and negatively correlated with membership of
190
N. Corrocher et al.
cluster 4: firms with a techno-organizational innovation mode tend to be smaller than the more conservative type. By looking more in depth, we observe that half of the firms in cluster 4 (53%) have one employee only, while this figure is well below 40% for firms in other clusters, particularly for firms in cluster 3 (27.2%). Cluster 1 exhibits the relatively highest share of large firms (3.5% as compared to a sample average of 2%), although the average size of its units is still rather small. This co-existence can be related to the fact that the innovative activity of firms in this cluster concentrates on the one side on service delivery, which requires a larger size to achieve economies of scale in distribution, and on the other side on tight cooperation with users/producers, which is peculiar of services provided by micro/individual firms. Our findings differ from those by Hollenstein (2003), who finds that the two most innovative clusters of services are characterized by large and medium size enterprises, and suggest a different structure of KIBS as compared to services as a whole. Our regression also shows that cluster membership is not significantly related to age, but in the case of interactive innovative modes (i.e. Cluster 1). Firms relying on novelties in service delivery and cooperation with other firms are likely to have been in the business longer than other types of KIBS. More than 50% of these firms were established for more than 15 years, as compared to a sample average of 41%. This is in line with the view suggesting that interactive learning requires time and experience. Therefore, it does not come as a surprise that firms need to be well-established in the market in order to benefit from cooperation with external actors. Firms in cluster 1 also tend to rely strongly on proximity, as they are less likely to find important customers at a distance, opening offices outside of the region. In this sense, the interactive mode of innovation exhibits some of the features which are typically associated with tertiary modes of production, such as co-localization with customers and intensity of interaction. Although particular significance is detected for cluster 1, markets are perceived to be rather local, defined by the same urban area or within a 50 km area, across all clusters (Fig. 1). The literature associates the local nature of markets with the intangible nature of services (de Jong et al. 2003), which makes longdistance relationships between suppliers and customers quite difficult. This intangible nature is particularly evident for some kind of professional services – precisely legal and accounting services - more than for computing services, the presence of which is more relevant in cluster 3. As Freel (2006) maintains, in the case of technology-intensive services (e.g. computing services), customers can easily defer to providers and do not need very close contacts, while when professional services are at stake, a more iterative approach is likely. The focus on highly localized markets is not exclusive. A non negligible share of firms finds important customers at the national level. On the other hand, foreign markets are not frequently targeted. Figure 1 shows that 44.4% of firms with foreign customers belong to cluster 4, due to the presence of a few very large consulting companies in the cluster. These firms are also likely to benefit more from the diffusion of ICT, which makes international markets more easily reachable.
Modes of innovation in knowledge-intensive business services. . . Fig. 1 Location of customers (%firms), by cluster
191
50,00 Cluster 1
40,00 Cluster 2
30,00
Cluster 3
20,00
Cluster 4
10,00 0,00 Same urban area
Within 50 km
Lombardy
Italy
Abroad
In terms of customer type, no significant difference is detected across clusters, as for all KIBS small and medium firms constitute the main market, accounting, respectively, for 29.7% and 30.4% of total clients. This evidence, which is consistent with the relevance of proximity to customers, does not come as a surprise within an economic system characterized by a highly fragmented and diversified manufacturing industry and by the prevalence of SMEs. Small firm innovation capacity relies extensively on external knowledge and the ability to absorb and to adapt to the internal idiosyncratic knowledge-base. The role of KIBS in supporting innovative SMEs is largely acknowledged in the literature (e.g. Muller and Zenker 2001), and the evidence that in Lombardy KIBS mainly interact with SMEs is consistent with this potential role of qualified interfaces for small units. SMEs and final consumers are extremely relevant customers for firms in cluster 4, although the difference with other clusters is not statistically significant. This is in line with the prevalence of specific professional KIBS in the techno-organizational innovation mode cluster - such as legal and accounting services - which typically find in individuals and households a relevant market, requiring a face-to-face contact and customized solutions. This evidence does not support the idea that KIBS are mostly specialist suppliers, which sell largely to other businesses and public institutions (Sundbo and Gallouj 2000; Miles 1995). In our case, the relevance of userproducer interaction extends also to cooperation with final consumers, and this is particularly true for professional KIBS providing legal and accounting services, as well as for architecture and engineering service firms. If we turn to the competitive environment, we observe that the probability of belonging to specific clusters is not affected by the intensity of competition or by competitor location. Competition is generally perceived to be rather diffused and increasing. More than 70% of the firms in the sample state that the number of direct competitors is greater than 10 and, most interestingly, this number is generally perceived as having substantially increased in recent times. Thirty-six percent of the firms believe more than ten new direct competitors entered their market over the last three years. Although all firms in the four clusters perceive a high degree of competition, this competition comes from
192
N. Corrocher et al.
different sources. In particular, firms in cluster 4 mostly face competition from very similar firms in terms of size: 60% of respondents in this cluster state that their main competitors have more or less the same number of employees as they have. On the contrary, firms in cluster 1 (where technical KIBS are overrepresented) perceive that they face competition mostly from larger firms. This is at odds with the literature that shows how competition from larger firms concerns mostly professional KIBS. However, it is worth noting that technical services are extremely heterogeneous: the evidence we find probably reflects the presence of engineering and architecture service firms in this cluster, which perceive competition from larger firms, as they focus not only on price/cost competition, but also on service quality and differentiation. In terms of competitor location, we observe that competition is strongly localized for most firms (Fig. 2). Around 70% of firms in each cluster state that their main competitor is located in the same urban area. The perception that the competitive game is played directly with other national firms outside the region is more diffused across technical KIBS providing computing services: cluster 3—which includes relatively more computing service firms—records the highest percentage of firms stating that their main competitor is an Italian firm or a foreign firm (11% as compared to a sample average of 8%). In general, however, international actors are rarely considered as direct competitors. The most striking differences across clusters are related to the identification of the most relevant factors for competitiveness. The competitive variable ‘Distribution channels’ negatively affects the probability of being in cluster 3 as compared to other clusters. More innovative modes are likely to imply a greater focus on distribution processes and post-sale services as competitive variables. ‘Innovativeness’ positively and significantly affect the probability of being in cluster 4 as compared to the more conservative cluster. Furthermore, techno-organizational and product innovation attitudes are positively associated with ‘Price’, a factor capturing also speed of service delivery and range of services offered. The variable ‘Reputation’ negatively affects the participation to cluster 1 (although the significance level is very low) the interactive mode of which seems to rely little on market wide visibility.
Fig. 2 Location of competitors (%firms), by cluster
80 Cluster 1
70
Cluster 2
60
Cluster 3
50
Cluster 4
40 30 20 10 0 Same urban area
Same region
Italy
Abroad
Modes of innovation in knowledge-intensive business services. . .
193
Finally, we investigate how cluster membership is affected by human resources management. In general, when asked about the employment of new professional profiles, KIBS reveal a rather conservative attitude: only 3% of firms in the sample state that they have hired new profiles over the last three years. No significant differences exist across clusters in this respect. Even if KIBS are not inclined to hire new personnel, they stimulate human resources to engage in training programs to update their competences. Training is especially relevant in cluster 2 and cluster 4, whose firms invest significantly more than the conservative ones. It is particularly in the case of focus on product innovation (Cluster 2) that we detect emphasis on training of human resources.
5 Conclusions The paper has investigated the existence of different innovation patterns across advanced services in Lombardy, which cut across the traditional dichotomy ‘technical vs. professional KIBS’. The empirical evidence in fact points at significant heterogeneity. First, results indicate that most but not all KIBS can be described as having a well-defined orientation towards innovation and change. Indeed, around 24% of KIBS in our sample do not appear to be particularly innovative and, as a consequence, rely upon well-established brand reputation to compete in the market, more than on other competitive variables. Furthermore, we find that, even within groups of innovative KIBS, innovation takes place in various forms, reflecting different strategies. In particular, our findings highlight four profiles of KIBS, which somehow empirically confirm the modes of innovation theorized by some of the existing studies (Tether 2005; Hollenstein 2003), but are related to a specific sub-sector, namely the KIBS, within the more general service sector. Besides a conservative mode, which identifies firms that do not carry out any relevant innovation activity, we find first a cluster conforming to a product innovation mode, which closely resembles the features of the manufacturing sector. Second, we identify a cluster characterized by features that are specific to the service sector, such as the interaction with other firms and customers (the interactive innovation mode). Finally, we have a mode of innovation which combines elements that are ascribed to the service sector by the traditional view of innovation in services (Pavitt 1984), such as the adoption of external technology, with other, softer sources of innovation, such as organizational change and investment in human capital. The techno-organizational mode suggests that technology adoption in KIBS is not an isolated and passive strategy, but is closely intertwined with changes associated with the way in which services are provided and organized, and in turn affects the relation with users. When examining more in depth the variables that are associated with cluster membership, we find that the firm competitive strategies in terms of distribution, price, reputation and innovativeness appear as the most significant determinants, with size, customer location, and training of human resources also playing a role in defining cluster
194
N. Corrocher et al.
specificities. This finding clearly supports some recent claims in the literature (see Tether 2005; Hollenstein 2003) that there is not a unique service pattern of innovation, but rather a variety of modes of innovation. However, it goes beyond that, since it takes into account one of the most innovation-oriented groups of services and investigates in depth firm heterogeneity in terms of innovation modes. It is important to stress that our results are inevitably affected by country and region specific factors, such as the composition of KIBS in terms of size (with a prevalence of micro firms), the peculiar competitive environment in which most professional and some technical KIBS operate - with regulatory restrictions hindering easy access to the market—the specific laws of the Italian labor market, which are likely to determine an underestimation of the actual number of human resources involved in some service activities. When considering this regional specificity and the representative character of the regional sample investigated, some policy implications can be advanced. Although KIBS are certainly key nodes of regional knowledge networks and providers of knowledge intensive inputs for local manufacturing, their strategies do not entirely and evenly reflect innovative attitudes. In addition, the evidence points at routine human resource upgrading, but lack of attention towards new competences and profiles. This is certainly a critical node in the current knowledge intensive tertiary transformation.
Acknowledgements The authors would like to acknowledge funding from Istituto Regionale di Ricerca della Lombardia (IRER), and to thank the participants at the 5th International EMAEE conference held in Manchester on 17–19 May 2007, and in the seminar held at Universitá del Piemonte Orientale (Novara) on 30 January 2008 for their useful comments and suggestions. Nicoletta Corrocher acknowledges the financial support of the Research Council of Norway (Project n˚172603/V10: “The Knowledge-based society”).
Appendix
Table 7 Factors for competitiveness
Availability of distribution channels Location of distribution channels Post-sale services Price competition Speed of service delivery Large range of services offered Collaboration with other firms Use of advanced technologies Development of new services Reputation Quality of services
Distribution channels
Price
Innovativeness
Reputation
0.87 0.85 0.46 0.10 0.19 0.09 0.16 −0.02 0.32 −0.05 0.04
0.07 0.05 0.14 0.79 0.63 0.58 −0.07 0.45 0.42 −0.07 0.16
0.07 0.08 0.09 −0.09 0.14 0.40 0.85 0.58 0.47 0.02 0.01
−0.02 0.03 0.00 −0.10 0.32 0.02 −0.05 0.14 0.06 0.78 0.76
Extraction method: principal component analysis. Rotation method: Varimax with Kaiser normalization. Percentage of variance explained is 59.1.
Modes of innovation in knowledge-intensive business services. . .
195
References Barras R (1990) Interactive innovation in financial and business services: the vanguard of the service revolution. Res Policy 19:215–237 Bresnahan T, Brynjolfsson E, Hitt LM (2002) Information technology, workplace organization and the demand for skilled labor: firm-level evidence. Q J Econ 67:339–376 Bryson JR, Monnoyer MC (2004) Understanding the relationship between services and innovation: the RESER review of the European service literature on innovation, 2002. Serv Ind J 24(1):205–222 Camacho J, Rodriguez M (2005) How innovative are services? An empirical analysis for Spain. Serv Ind J 25(2):253–271 Coombs R, Miles I (2000) Innovation, measurement and services: the new problematique. In: Metcalfe JS, Miles I (eds) Innovation systems in the service economy - measurement and case study analysis. Kluwer Academic, Boston Czarnitzki D, Spielkamp A (2000) Business services in Germany: bridges for innovation. ZEW discussion paper, No. 00–52. Mannheim Doloreux D, Muller E (2007) The key dimensions of knowledge-intensive business services (KIBS) analysis. A decade of evolution, Working Paper Firms and Regions No. U1/2007, Fraunhofer-Institut für System-und Innovationsforschung-ISI, Karlsruhe Evangelista R, Savona M (2003) Innovation, employment and skills in services. Firm and sectoral evidence. Struct Chang Econ Dyn 14:449–474 Freel M (2006) Patterns of technological innovation in knowledge-intensive business services. Ind Innov 13(3):335–358 Gadrey J, Gallouj F (1998) The provider-customer interface in business and professional services. Serv Ind J 18(2):1–15 Gallouj F (2002) Knowledge intensive business services: processing knowledge and producing innovation. In: Gadrey J, Gallouj F (eds) Productivity, innovation and knowledge in services. Edward Elgar Gallouj F, Weinstein O (1997) Innovation in services. Res Policy 26:537–556 den Hertog P (2000) Knowledge intensive business services as co-producers of innovation. Int J Innov Manag 4(4):491–528 Hipp C, Tether B, Miles I (2000) The incidence and effects of innovation in services: evidence from Germany. Int J Innov Manag 4:417–453 Hollenstein H (2003) Innovation modes in the Swiss service sector: a cluster analysis based on firm-level data. Res Policy 32:845–863 Howells J (2000) The nature of innovation in services. Report presented to the OECD “Innovation and Productivity in Services Workshop”, Sidney, Australia. (http://www. oecd.org/dsti/sti/industry/indcomp) Howells J, Tether B (2004) Innovation in services: issues at stake and trends. European Commission, Contract No. INNO-03-01 Jensen MB, Johnson B, Lorenz E, Lundvall BA (2007) Forms of knowledge and modes of innovation. Res Policy 36:680–693 de Jong JPJ, Bruins A, Dolfsma W, Meijaard J (2003) Innovation in service firms explored: what, how and why?. Strategic Study B200205, EIM Business and Policy Research (http://www.ondernemerschap.nl/pdf-ez/B200205.pdf) Koschatzky K, Zenker A (1999) The regional embeddedness of small manufacturing and service firms: regional networking as knowledge source for innovation. Working papers firms and region R2/1999. ISI, Karlsruhe Kuusisto J, Viljamaa A (2004) Knowledge intensive business services ad co-production of knowledge—the role of public sector? Frontiers of E-Business Research, Conference proceedings of eBRF 2004. Tampere University of Technology and University of Tampere Larsen JN (2001) Knowledge, human resources and social practice: the knowledge-intensive business service firm as a distributed knowledge system. Serv Ind J 21(1):81–102 Leiponen A (2001) Knowledge services in the innovation system. Helsinki: Taloustieto Licht G, Moch D (1999) Innovation and information technology in services. Can J Econ - Revue Canadienne d’Economique 32:363–383 Miles I (1995) Services innovation: statistical and conceptual issues. PREST Working Paper, PREST, University of Manchester
196
N. Corrocher et al.
Miles I (2005) Knowledge intensive business services: prospects and policies. Foresight 7(6):39–63 Miles I, Kastrinos N, Flanagan K, Bilderbee R, den Hertog P, Huitink W, Bouman M (1995) Knowledge intensive business services: their role as users, carriers and sources of innovation, EIMS Publication No 15, Innovation Programme, DGXIII, Luxembourg Muller E, Zenker A (2001) Business services as actors of knowledge transformation: the role of KIBS in regional and national innovation systems. Res Policy 30:1501–1516 Nählinder J (2002) Innovation in knowledge intensive business services: state of the art and conceptualisation, Tema T WP 244. Linköping Pavitt K (1984) Sectoral patterns of technical change: towards a taxonomy and a theory. Res Policy 13:343–73 Simmie J, Strambach S (2005) The contribution of KIBS to innovation in cities: an evolutionary and institutional perspective. J Knowl Manag 10(5):26–40 Sirilli G, Evangelista R (1998) Technological innovation in services and manufacturing: results from Italian surveys. Res Policy 27:881–899 Strambach S (1998) Knowledge-intensive business services (KIBS) as an element of learning regions—the case of Baden Württenberg. Paper presented to the 38th Congress of the European Regional Science Association, Vienna Sundbo J, Gallouj F (2000) Innovation as a loosely coupled system in services. In: Metcalfe JS, Miles I (eds) Innovation systems in the service economy—measurement and case study analysis. Kluwer Academic Publishers, Boston Toivonen M (2004) Expertise as business: long-term development and future prospects of knowledge-intensive business services. Doctoral Dissertation Series, Helsinki University of Technology. Available at: http://lib.tkk.fi/Diss/2004/isbn9512273152/ Thomi W, Böhn (2003) Knowledge intensive business services in regional systems of innovation— initial results from the case of Southeast-Finland. 43rd European congress of the regional science association Tether BS (2003) The sources and aims of innovation in services: variety between and within sectors. Econ Innov New Technol 12(6):481–505 Tether BS (2005) Do services innovate (differently)? Insights from the European innobarometer survey. Ind Innov 12(2):153–184 Tether BS, Hipp C (2002) Knowledge intensive technical and other services: patterns of competitiveness and innovation compared. Technol Anal Strategic Manag 14(2):163–182 Tether BS, Hipp C, Miles I (2001) Standardisation and particularisation in services: evidence from Germany. Res Policy 30:1115–1138 Tether BS, Metcalfe JS (2004) Services and systems of innovation. In: Malerba F (ed) Sectoral systems of innovation. Cambridge University Press, Cambridge Wood PA, Bryson J, Keeble D (1993) Regional patterns of small firm development in the business services: evidence from the UK. Environ Plan A 25:677–700