The location effect on KIBS firm innovation capacities: An empirical study Cristina I. Fernandes Polytechnic Institute of Bragança (IPB); Instituto Superior de Línguas and Administração de Leiria (ISLA), and NECE - Research Unit in Business Sciences, Covilhã, Portugal Email: [email protected] João J. M. Ferreira University of Beira Interior (UBI) and NECE - Research Unit in Business Sciences, Covilhã, Portugal Email: [email protected]

Abstract This research seeks to investigate just how location impacts on the innovation capacities of knowledge intensive business service (KIBS) firms. Does locating near the main urban centres influence the capacity of these firms to leverage both innovation and financial performance? In this article, we find that location does hold an influence over KIBS firm innovation capacities and it is therefore correspondingly relevant to take firm location into consideration and thereby contributing to the literature on innovation. The results of the econometric models estimated for a sample of 442 KIBS demonstrate location is a variable influencing firm innovation capacity. The closer the location to the larger urban centres, the greater the innovative capacity and financial performance of this firm type. Different levels and types of innovation were also identified by firm location.

Key Words: Location, Innovation, KIBS, Performance.

1. Introduction Ever since the 1980s, there has been steadily rising interest in the service sector, which gained international recognition and especially within the framework of regional development as Europe and North America grew concerned over the fallout from

1

deindustrialisation (Muller & Doloreaux, 2009). Hitherto, services had been broadly dismissed as mere subsidiaries of the transformative activities. Furthermore, in the last two decades, attention has particularly focused on the functions served by the service sector. Despite the growing awareness that innovation is not limited to technical processes and products, recent research has mostly concentrated exclusively on technical innovation and especially in the transformative industrial sector (Becker & Dietz, 2004; Huergo & Jaumandreu, 2004; Lynskey, 2004; Nieto & Santamaria, 2005). Only more recently has attention turned to focus on innovation in the service sector (Gallouj & Weinstein, 1997; Tether, 2003). According to Tether et al. (2001), innovation in service firms is traditionally perceived as something taking place only slowly and over the course of much time. Services were seen as incapable of innovation in and of themselves and only adopted innovations generated by transformative industry firms. Additionally, according to Pavitt (1984), the smaller the scale of service firms, the less they tend to directly engage in research and development (R&D) proving mere recipient accumulators of technology and innovation produced by other sectors. Within the service industry, the rapid growth of the knowledge intensive business service (KIBS) sector has reflected its increasingly important role in innovation processes (Muller, 2001; Howells & Tether, 2004; Toivonen, 2004; Koch & Stahlecker, 2006). This KIBS involvement in innovation is affirmed above all by the fact that these firms play a complex role (and not simply meeting the needs of market demand but more specifically those of its clients) involving establishing bridges of knowledge and innovation between firms and science (Miles et al., 1995; Czarnitzki & Spielkamp, 2003). This emphasis on KIBS firms rises to the extent that some authors propose we are on the cusp of a third industrial revolution (Tether & Hipp, 2002). Sheamur and Doloreaux (2008) demonstrate how KIBS contribute to innovation and regional competitiveness through the way in which they interact with other local actors in the production of innovation and consequently boosting regional development through generating synergies. According to Cooke et al. (2004), firm innovation capacities and strategies vary in accordance with the region where they are located (Cooke et al., 2004). Thus, we arrive at the following research questions: where do KIBS firms opt to locate and what are the effects of this location choice on their innovative capacity? This research project analysed these issues in order to provide a better understanding of the effects of location on this firm type.

2

The locations and contributions to local economies of these firms have been subject to analysis by diverse authors (O hUallachain & Reid, 1991; Coffey & Shearmur, 1997; Gong, 2001). Their location in urban environments, their sensitivity to general economic agglomeration (Eberts & Randall, 1998; Poehling, 1999; Wernerheim & Sharpe, 2003) and their tendency to form spatial clusters (Coe, 1998; Keeble & Nachum, 2002) has been documented through recourse to various different methodologies. A large majority of these studies seek to research local economic dynamics, regional development and the reasons behind some regions growing faster and further than others (Moyart, 2005). Despite the growing body of research on this field, there are still few studies examining the impact of KIBS innovative capacities on regions (Shearmur, 2011). Particularly in relation to the KIBS firm location, earlier research has returned contradictory results. According to Malecki et al. (2004), KIBS firms mostly locate in cities as these are the sites par excellence not only for business sector innovation but also for the networks able to drive such innovation. However, Sheamur and Doloreaux (2008) find that KIBS actually prefer to opt for rural locations. These findings reflect a need for further research concentrating on the location of this firm type given the importance of location both to the academic field in itself and to business practice in general. There is much yet to be appropriately explained and we still do not understand the way in which business managers take decisions on location. In this article, we propose that location does wield an effect on the innovation capacities of KIBS firms and therefore confirming that the location of these firms does play a relevant role and correspondingly overcoming a gap thus far existing in the literature. Hence, the objective of this study is that of exploring the effects of location: does proximity to major urban centres influence the innovative capacities of KIBS firms and consequently raise their financial performance levels? The article is structured as follows: following this introduction, section two approaches firm innovative capacities and the relationship between those present at KIBS firms and their respective locations. In section three, we set out both the sample as well as the methodology applied. Section four then presents and discusses the results before we close with our main conclusions.

2. Literature Review

3

Innovation is a process involving the organisation in its full extent while simultaneously conditioning organisational behaviour (Kline, 1985; Nelson & Winter, 1982; Wakelin, 1998; Yam et al., 2011). Study of the organisational variables conditioning innovation opens up a very important path to grasping the innovation capacities in effect at firms (Román et al., 2011). The innovative behaviour of firms is commonly measured according to their innovation capacities (Cohen & Levinthal, 1990; Teece et al., 1997). The firm has to internally adapt to whatever needs stem from the innovation process in order to attain the sought after innovative capacity (Nonaka & Takeuchi, 1995; Hult et al, 2004). Capacity is thus fundamental to studying innovation, which in turn raises the question: just what is innovative capacity? According to Hull & Covin (2010), the greater the capacity for learning, the quicker the firm will be able to respond and meet changing marketplace demands, such as through the launch of new products. Hence, the greater the intensity of knowledge, the greater the innovative capacity level. Forsman (2011) proposes that innovative capacity is a phenomenon made up of variables including the firm’s internal resources and capacities and its participation in networks of cooperation. Innovation is, in every sector of the economy, fundamental for firms surviving and prevailing in an increasingly globalised world. Innovation enables firms to respond to the diversified and constantly evolving demands and ensures improvements are leveraged across the different domains and activities of society (Cooke et al, 2004). Thus, innovation is seen as the motor of progress, of competitiveness and of economic development (Romer, 1994; Johansson et al., 2001). Correspondingly, innovative firms may be expected to turn in better economic-financial performances (Ferreira, 2010; Kostopoulos et al., 2011; Forsman, 2011). Currently, there is support for the idea that there are differences in the innovation performances registered between regions with the respective firm innovation levels and strategies conditioned by the surrounding region of location (Cooke et al., 2004). However, government support policies also prove fundamental to innovation capacities and technological changes in regions, and especially in rural areas (Doloreux & Dionne, 2008). According to the OECD (2007), the motivation that should underpin studies on the differences between regional innovation differentials is the identification and justification for policies enabling more disadvantaged regions to attain better innovation performance levels. Nevertheless, there are still few studies examining the impact of innovation brought about by KIBS firms at the regional level (Shearmur & Doloreux, 2008). The KIBS studies thus far published focus analysis on: (i) the impact on

4

employment (Chadwick & Glasson, 2008; Shearmur & Doloreux, 2008); (ii) the impact on the growth of cities (Simmie & Strambach, 2006; Aslesen & Isaksen, 2007a); (iii) the effect of proximity on the transfer of knowledge and on clients (Aslesen & Jakobsen, 2007); (iv) the effect of innovation on the transformative industry and the relationship with cooperation (Muller, 2001; Aslesen & Isaksen, 2007b; Muller & Doloreux, 2009); and (v) the relationship with the innovation systems in effect in the surrounding environment (Hu et al., 2006; Koch & Stahlecker, 2006). Miles et al. (1995) distinguish three key KIBS characteristics: (i) the high importance such firms attribute to professional knowledge; (ii) their desire to be the entities making primary recourse to information and knowledge or deploying their knowledge to produce services and acting as intermediaries between these services, their clients and their production processes; and, (iii) the great importance of the service types supplied to firms in terms of competition and competitiveness. At this point, we should emphasise that the literature breaks KIBS firms down into two main categories (Frell, 2006; Miles et al., 1995; Doloreux & Muller, 2007, Shearmur & Doloreux, 2008): (1) technological KIBS (t_KIBS), which covers activities related to information technologies, R&D, engineering, architecture, consulting, testing and analytical techniques; and (2) professional KIBS (p_KIBS), incorporating firms providing legal, accountancy, registration, auditing, fiscal consultancy and market study services as well as the entire publicity sector. According to research carried out by Frell (2006), the t_KIBS employ members of employees with higher levels of qualification with this interrelated with their levels of innovation whilst in the case of p_KIBS, suppliers and clients drive innovation. KIBS are, irrespective of their type, perceived as directly influencing the innovation capacities of their clients and thereby making a fundamental contribution to innovation systems. Cooke and Leydesdorff (2006) back this position in arguing that KIBS represent an important link in the creation of local innovation infrastructures, thus contributing to the region building up its own competitive advantage. Similarly, Aslessen and Isaksen (2007) verify a strong KIBS presence in urban areas and basing their explanation for this on the laws of supply and demand. Urban environments have more favourable pre-conditions, such as knowledge producing organisations (universities and research centres) able to contribute to developing KIBS activities (Keeble & Nachum, 2002; Aslesen & Isaksen, 2007).

5

Other researchers advocate studying the innovation actually ongoing at KIBS firms and not in terms of the relationship with other firms and thereby making their respective surrounding region more competitive in relation to others. KIBS correspondingly tend to prefer locating in regions displaying socio-cultural and institutional structures enhancing the incidence of continuous learning and innovation, which reflects in a preference for entrepreneurial regions (Markusen, 1999; Keeble & Nachum, 2002; Cooke et al., 2004; Doloreux, 2004). KIBS are furthermore identified in regions where the need for development becomes more necessary (McCann, 2007). Therefore, different firm types require different types of knowledge intensity and contact networks and KIBS firms correspondingly tend to locate more in urban areas where such networks may be more easily leveraged thereby facilitating the sharing of intensive knowledge (Crevoiser & Camagni, 2001; Malecki, 2007).

3. METHODOLOGY 3.1. Data The sample was obtained from a database made up of a total of 34,971 KIBS firms. We selected firms according to their CAE (Rev.3) and NACE (Rev. 2) codes, in keeping with other research projects (Frell, 2006, Miles et al., 1995; Doloreux & Muller, 2007, Shearmur & Doloreux, 2008), in order to include both KIBS firm types: t_KIBS (which covers activities related to information technologies, R&D, engineering, architecture, consulting, testing and analytical techniques), and p_KIBS (incorporating firms providing legal, accountancy, registration, auditing, fiscal consultancy and market study services as well as the entire publicity sector). Either a face-to-face and/or via telephone questionnaire was completed at 442 KIBS firms on mainland Portugal. The sample was selected by the convenience method and included all 18 Portuguese districts in the research. We correspondingly carried out approximately 25 questionnaires in each NUT III region, of which 27.8% belong to the North Region, 35.7% to the Centre Region, 7.5% to the Lisbon Region, 23.3% to the Alentejo Region and 5.7% to the Algarve Region.

3.2. Definition and measurement of variables

6

In terms of the methodology utilised to describe the KIBS profiles, we determined the absolute and relative frequencies for the qualitative variables whilst calculating averages and standard deviations for the quantitative characteristics. As regards our modeling of the factors of location influencing innovations, we adopted two approaches: one took into account the firm location in terms of its NUT II region and the other accounted for the distance to the main Portuguese urban conurbations, Lisbon and Oporto. As regards innovations, we studied the number of service innovations, the number of process innovations, and the total number of innovations (services, processes, organisational, introduction of already existing services into new markets, patents, launching brands and new designs for services and processes). In the models calculated, the dependent variable corresponds to a discrete count variable in accordance with the count model approach. We ran calculations according to the Poisson and Negative Binomial models in addition to the Akaike Information Criterian (AIC) and Bayesian Information Criterian (BIC) and the R2. The method best adjusting to the data is based upon Poisson distribution and hence this methodology was correspondingly adopted (Cameron & Trivedi, 2005). The Poisson distribution for determining the probability of events occurring over a period time is calculated by , in which µ corresponds to the average process rate. Poisson

regression is derived from the Poisson distribution through a re-parameterisation of the relationship between the average µ and the regressors x (

.

In this model, the control variables introduced to the regressions were the ages of the firm and its manager/owner, KIBS type (p_KIBS vs t_KIBS) and business turnover (less to €50,000 vs. over €500,000 annually). The Poisson regression parameters were estimated by the maximum likelihood method with robust standard levels of estimated error eliminating heteroscedasticity given this regression methodology is intrinsically heteroscedastic, introducing a bias into the estimate variance compromising the validity of the statistical significance of estimates. We estimated three models for each of the aforementioned innovations. For model I, the independent variables, in addition to firm characteristics, with location in terms of NUT II applied to serve as a dummy variable. Model II applied the shortest distance to Lisbon

and Oporto

(more specifically

,

and the square

of these variables) as their independent variables. In model III, we deployed the regional

7

dummies and the distances. To evaluate the model quality of adjustment, we adopted the pseudo-R2 proposed by Cragg and Uhler and applied the Vuong test for comparison. All the calculations made recourse to the pscl procedure in software R version 2.14 (R Foundation for Statistical Computing) and IBM-SPSS version 18.0 (IBM Corporation). Table 1 provides the results returned by the variables characterising the firms. The results are presented both for the sample as a whole and broken down by NUT II location.

3.2.1 KIBS type and location The distinction between KIBS type, whether professional (p_KIBS) or technological (t_KIBS) based, is identified by various authors (Frell, 2006; Miles et al., 1995; Doloreux & Muller, 2007, Shearmur & Doloreux, 2008). This sample includes 34.2% t_KIBS and 65.8% p_KIBS (table 1).

3.2.2

Financial performance and location

There is broad backing from a range of authors that innovation capacities positively influence financial performance (Ferreira, 2010; Kostopoulos et al., 2011; Forsman, 2011). In our study, we sought to measure this impact in detail. Financial performance is measured through turnover in keeping with other studies (He & Wong, 2004; Fusfuri & Tribó, 2008) We found that 34.4% turned over less than €50,000 (33% in p_KIBS and 37.1% in t_KIBS), 25.2% between €50,000 and €100,000 (28.9% in p_KIBS and 17.9% in t_KIBS), 19.9% between €100,00 and €200,000 (19.6% in p_KIBS and 17.9% in t_KIBS) and 7% over €500,000 (4.8% in p_KIBS and 11.3% in t_KIBS) (table 1).

3.2.3

Profile of the firm leader and location

The need to establish a relationship between the decisions of firm leaders and their personal characteristics, such as their parents’ professions, gender (male/female), ethnic background, academic qualifications, years of experience in the sector of activity; age and firm years of operation have all attracted the attention of researchers (Mitchell et al., 2002; Lafuente et al., 2010). The average reported age of firms was 11±6.9 years (11.6±7 years at p_KIBS and 9.9±6.5 years at t_KIBS) and their managers was 41.9±8.1 years (42.4±8.2 years at p_KIBS and 41±8 years at t_KIBS), and higher in the Lisbon region (43.3±8.5).

8

3.2.4

Characterisation of human resources and location

Various authors point out how organisations employing better qualified human resources display a greater ability to engage in innovation processes (Oakey et al., 1980; Jaffe et al, 1993; Paci & Usai, 2000; Kim & Knaap, 2001; Sohn et al., 2003; Fischer & Varga, 2003). Miles et al. (1995) and Delmar and Wennberg (2010) state that as KIBS firms deploy and develop highly intensive knowledge then they also employ members of employees with higher education levels of academic qualifications. According to Frell (2006), t_KIBS display a need for more intensive knowledge and corresponding employ more highly qualified professionals than p_KIBS. In table 1, we find that the average number of employees stands at 4.5±7.5 (in the North this drops to 3.7±2.6 while rising to 6.1±13.7 in the Alentejo, and 4.6±8.7 years at p_KIBS and 4.3±4.2 years at t_KIBS), with 81.2% of employees on average having completed university level education (in the Algarve, this percentage dropped to 72.5%, and stood at 81.4% at p_KIBS and 27.4% at t_KIBS).

9

Table 1 – Descriptive statistics: profile of business leader by KIBS location Region North

Centre

KIBS

Lisbon

Alentejo

Algarve

Total

p_KIBS

N

%

N

%

N

%

N

%

N

%

N

%

p_KIBS

78

63.40%

107

67.70%

19

57.60%

69

67.00%

18

72.00%

291

65.80%

t_KIBS

45

36.60%

51

32.30%

14

42.40%

34

33.00%

7

28.00%

151

34.20%

Less than €50,000

42

34.10%

54

34.20%

8

24.20%

40

38.80%

8

32.00%

152

33

26.80%

44

27.80%

4

12.10%

27

26.20%

3

12.00%

21

17.10%

28

17.70%

13

39.40%

18

17.50%

8

10

8.10%

9

5.70%

2

6.10%

5

4.90%

7

5.70%

5

3.20%

1

3.00%

5

2

1.60%

4

2.50%

1

3.00%

8

6.50%

14

8.90%

4

12.10%

t_KIBS

N

%

N

%

34.40%

96

33.00%

56

37.10%

111

25.10%

84

28.90%

27

17.90%

32.00%

88

19.90%

57

19.60%

31

20.50%

4

16.00%

30

6.80%

17

5.80%

13

8.60%

4.90%

1

4.00%

19

4.30%

15

5.20%

4

2.60%

4

3.90%

0

0.00%

11

2.50%

8

2.70%

3

2.00%

4

3.90%

1

4.00%

31

7.00%

14

4.80%

17

11.30%

KIBS

From €50,000 €100,000 From €100,000 €200,000 Financial From €200,000 Performance €300,000 From €300,000 €400,000 From €400,000 €500,000 Over €500,000

to to to to to

Mean ± SD

10.8 ± 7.1

10.7 ± 7.4

11.2 ± 6.4

11.5 ± 5.7

11 ± 7.2

11 ± 6.9

11.6 + 7

9.9 + 6.5

Median (Min - Max)

10 (1 - 46)

10 (1 - 29)

9 (3 - 29)

11 (1 - 27)

11 (1 - 27)

10 (1 - 46)

11 (1 - 46)

9 (1 - 29)

Mean ± SD

41.4 ± 8.3

42 ± 8.6

43.3 ± 8.5

41.8 ± 6.9

41.7 ± 8.9

41.9 ± 8.1

42.4 + 8.2

41 + 8.0

40 (27 - 73)

41 (24 - 68)

42 (30 - 60)

41 (28 - 59)

41 (30 - 61)

41 (24 - 73)

41 (24 - 73)

40 (27 - 65)

4.4 ± 5.5

3.7 ± 2.6

4.6 ± 4

6.1 ± 13.7

4 ± 2.3

4.5 ± 7.5

4.6 + 8.7

4.3 + 4.2

3 (1 - 50)

3 (1 - 21)

3 (1 - 17)

4 (1 - 110)

4 (1 - 10)

3 (1 - 110)

3 (1 - 110)

4 (1 - 44)

78.4% ± 30.7%

84% ± 22.2%

82.9% ± 18.6%

81.6% ± 27.3%

72.5% ± 29.8%

81.2% ± 26.3%

81.4% ± 25.8%

80.8% ± 27.4%

100% (0% - 100%)

100% (0% - 100%)

83% (33% - 100%)

100% (0% - 100%)

80% (0% - 100%)

100% (0% - 100%)

100% (0% - 100%)

100% (0% - 100%)

Firm Age

Manager Age

Median (Min - Max)

Total Mean ± SD number of Median (Min - Max) employees % Graduate employees

Mean ± SD Median (Min - Max)

Note: SD – Standard deviation; Min – Minimum; Max - Maximum

10

3.2.5

Innovative capacities

Sundbo (1998) advocates the measurement of service sector innovation according to: (i) new products or services; (ii) new processes; (iii) new forms of organisation and management; (iv) new marketing techniques; (v) alterations to the physical form of objects; (vi) changes in intellectual terms (consultancy services); (vii) new means of transporting products; (viii) launching new strategies. Furthermore, according to Camacho and Rodrigues (2005), this requires adopting a combination of theories, ranging from the most recent to the most longstanding studies on service sector innovation. Indeed, innovation in this type of sector needs grasping beyond the mere introduction of new products or processes. Furthermore, the growing importance of firm innovation, especially KIBS, to competitiveness and regional development has gained growing recognition in the literature (Malecki et al, 2004; Wood, 2005; Muller & Doloreux, 2009). The literature also backs the significance of the role of KIBS in regional innovation systems, especially for transformative industry support activities and small and medium sized firms in general (Cooke, 2001; Wood, 2005). Table 2 describes the number of innovations, by typology and according to NUT II. On average, the sample firms undertook 1.7±0.9 innovations in services in 2009, with the Alentejo the region with the lowest average number of innovations (1.5±0.8). In terms of the average number of process innovations, this returned a lower rate (0.1±0.4) with the total number of innovations averaging out at 1.8± 1. Table 2 – Descriptive statistics: number of innovations by locations Region North

Centre

Lisbon

Alentejo

Algarve

Total

Innovations services

in Mean ± DP Median (Min - Max)

1.8 ± 0.9

1.7 ± 0.9

2.1 ± 0.9

1.5 ± 0.8

1.9 ± 0.8

1.7 ± 0.9

2 (0-3)

1 (0-4)

2 (1-3)

1 (0-4)

2 (1-3)

2 (0-4)

Innovations processes

in Mean ± DP Median (Min - Max)

0.1 ± 0.6

0 ± 0.3

0±0

0.1 ± 0.3

0 ± 0.2

0.1 ± 0.4

0 (0-4)

0 (0-3)

0 (00)

0 (0-2)

0 (0-1)

0 (0-4)

Total innovations

of Mean ± DP

1.9 ± 1

1.9 ± 1.2

2.1 ± 0.9

1.6 ± 0.9

1.9 ± 0.8

1.8 ± 1

2 (0-5)

2 (1-9)

2 (1-3)

1 (0-6)

2 (1-3)

2 (0-9)

Median (Min - Max)

Note: SD – Standard deviation; Min – Minimum; Max - Maximum

11

4. RESULTS 4.1. Econometric estimations

The econometric model estimations for the number of service innovations are set out in Table 3. Model 1, including the variables of firm location by NUT II region, reports that regarding the control variables, firm age (B=-0.02, Z=-3.82, p < .01) and manager age (B=0.02, Z=3.26, p < .01), have a statistically significant influence on the number of innovations in services. Hence, the older the firm, the fewer the number of innovations and the older its manager, the greater the number of innovations. In terms of the regional dummy variables, we report that firms in the North (B=0.20, Z=3.08, p < .01), Centre (B=0.13, Z=2.11, p <.05), Lisbon (B=0.31, Z=3.42, p < .01) and the Algarve (B=0.26, Z=2.73, p < .01) produce, on average, a greater number of service innovations than firms located in the Alentejo. In model 2, which includes base 10 location logarithm variables for the distances to Lisbon and to Oporto, as well as their squared results, we also find that firm age (B=0.02, Z=-4.09, p < .01) and manager age (B=0.02, Z=3.50, p < .01) has a statistically significant influence on the incidence of service innovations. As regards the distance related variables, there is no statistically significant influence on the number of service innovations. In model 3, including the regional dummy variables and the distances, we find that firm age (B=-0.02, Z=-2.99, p < .01) and manager age (B=0.02, Z=3.39, p < .01) has a statistically significant influence on the incidence of service innovations. As regards the location variables, we find that firms in the North (B=0.43, Z=3.21, p < .01), Centre (B=0.25, Z=2.85, p < .01), Lisbon (B=0.35, Z=2.14, p < .05) have higher rates of service innovations and the greater the distance from Oporto (B=0.26, Z=2.56, p < .05), the greater the number of innovations. The Vuong test reports that Model 3 is statistically more explicative at p < .10 than Model 1 and at p < .01 is more explicative than Model 2. Thus, we find Model 3 is significantly better able to explain the number of service innovations taking place at KIBS firms.

12

Table 3 – Count models: Dependent variable – Number of service innovations Model 1

Model 2

Model 3

B

SE

Z

B

SE

Z

B

SE

Z

Intercept

-0.10

0.18

-0.56

0.02

0.19

0.13

-0.34

0.20

-1.70

t_KIBS

0.02

0.05

0.45

0.02

0.05

0.48

0.02

0.05

0.47

Firm age

-0.02

0.01

-3.82***

-0.02

0.01

-4.09***

-0.02

0.01

-3.86***

Manager age

0.02

0.00

3.26***

0.02

0.00

3.50***

0.02

0.00

3.53***

TO > €400,000

0.13

0.09

1.49

0.14

0.09

1.65*

0.12

0.09

1.38

No. employees

0.00

0.00

0.26

0.00

0.00

-0.23

0.00

0.00

0.25

Graduate employees (%)

0.06

0.09

0.68

0.05

0.09

0.49

0.10

0.09

1.10

North

0.20

0.07

3.08***

0.43

0.13

3.21***

Lisbon

0.31

0.09

3.42***

0.35

0.16

2.14**

Centre

0.13

0.06

2.11**

0.25

0.09

2.85***

Algarve

0.26

0.09

2.73***

0.20

0.11

1.86*

log10(DL) log10(DP)

0.00

0.09

0.00

-0.03

0.17

-0.18

0.02

0.06

0.28

0.26

0.10

2.56**

2

(log10(DL))

0.04

0.06

0.65

-0.04

0.08

-0.53

(log10(DP))2

0.00

0.06

-0.03

0.11

0.08

1.42

N

437

437

437

Pseudo-R2

0.071

0.058

0.077

-2LL

30.18

24.59

32.99

AIC

1227.3

1232.9

1232.5

BIC

1272.2

1277.8

1293.7

V test (M1 vs M2)

1.93**

V test (M1 vs M3)

-1.32*

V test (M2 vs M3)

-2.33***

Note: * p < .10, ** p < .05, *** p < .01; B: Regression coefficient; SE: Standard error; Z: Z-test; TO > 400.000€ - Turnover greater than €400,000; Graduate employees (%):Percentage of employees holding university level education qualifications; DL: Distance to Lisbon; DP: Distance to Oporto; Pseudo-R2: Cragg and Uhler's pseudo r-squared; -2LL: -2 Log likelihood; AIC: Akaike Information Criterion; BIC: Bayesian Information Criterion; V Test: Vuong test

13

The process innovation models are set out in Table 4. In model 1, which includes the firm location by NUT II region variable, we find that the control variables, firm age (B=-0.10, Z=-2.17, p < .05) and manager age (B=0.14, Z=2.75, p < .01) generate a statistically significant influence on the number of service innovations. As regards the dummy regional variable, the sample reports that firms located in the Lisbon region (B=-15.67, Z=-25.93, p < .01) on average implement a lower number of process innovations. In model 2, only manager age (B=0.13, Z=2.42, p < .05) proves to have any statistically significant influence on the number of process innovations. As regards the variables for distances, they display no statistically significant influence over process innovations. In model 3, we find manager age (B=0.13, Z=2.35, p < .05) has a statistically significant influence on the number of process innovations. In the case of the location variables, the findings show how the greater the distance from Oporto (B=-1.38, Z=-2.00, p < .05), the lesser the number of process innovations. The Vuong test reports that Model 3 and Model 2 are statistically equal in explanatory terms terms and are both statistically more explicative than Model 1 at p < .05.

14

Table 4 – Count Models: Dependent variable – Number of process innovations Model 1

Model 2

Model 3

B

SE

Z

B

SE

Z

B

SE

Z

Intercept

-8.08

2.17

-3.72

-8.82

2.14

-4.13

-7.50

2.00

-3.75

t_KIBS

-0.16

0.53

-0.31

0.29

0.47

0.62

0.33

0.39

0.84

Firm age

-0.10

0.05

-2.17**

-0.06

0.04

-1.53

-0.06

0.04

-1.56

Manager age

0.14

0.05

2.75***

0.13

0.05

2.42**

0.13

0.06

2.35**

TO > €400,000

0.34

0.76

0.45

-0.01

0.44

-0.01

0.26

0.65

0.40

No. employees

-0.03

0.04

-0.80

-0.04

0.05

-0.89

-0.06

0.07

-0.95

Graduate employee (%)

0.55

0.82

0.67

0.36

1.09

0.33

0.34

1.13

0.30

North Lisbon

0.79 -15.67

0.63 0.60

1.25 -25.93***

-1.61 -32.83

1.21 39.75

-1.33 -0.83

Centre

-0.74

0.86

-0.86

-1.22

1.04

-1.18

Algarve

-0.39

1.14

-0.34

-0.22

1.22

-0.18

log10(DL) log10(DP) 2

(log10(DL))

2

(log10(DP))

2.58

2.05

1.26

-1.29

4.99

-0.26

-0.56

0.59

-0.96

-1.38

0.69

-2.00**

-2.45

2.55

-0.96

3.94

7.57

0.52

0.49

0.32

1.53

0.20

0.40

0.50

N Pseudo-R2

437 0.181

437 0.313

437 0.332

- 2 LL

35.59

63.66

67.83

AIC

234.1

206.0

209.82

BIC

278.9

250.9

271.0

V test (M1 vs M2)

-1.77**

V test (M1 vs M3)

-2.02**

V test (M2 vs M3)

-0.61

Note: * p < .10, ** p < .05, *** p < .01; B: Regression coefficient; SE: Standard error; Z: Z-test; TO > €400,000 – Turnover greater than €400,000; Graduate employees (%): Percentage of employees holding university level education qualifications; DL: Distance to Lisbon; DP: Distance to Oporto; Pseudo-R2: Cragg and Uhler's pseudo r-squared; -2LL: -2 Log likelihood; AIC: Akaike Information Criterion; BIC: Bayesian Information Criterion; V Test: Vuong test

15

Finally, we modelled the total number of innovations and these results are detailed in Table 5. Model 1 reports that firm age (B=-0.02, Z=-3.49, p < .01), manager age (B=0.02, Z=3.51, p < .01) and annual firm turnover in excess of €400,000 (B=0.24, Z=2.33, p < .05) are factors with a statistically significant influence on the number of innovations. As regards the regional dummy variable, this demonstrates how firms in the North (B=0.20, Z=2.88, p < .01) and Lisbon (B=0.21, Z=2.26, p < .05) regions undertake a higher average number of innovations. In model 2, we also observe firm age (B=-0.02, Z=-3.32, p < .01), manager age (B=0.02, Z=3.45, p < .01) and annual firm turnover in excess of €400,000 (B=0.25, Z=2.40, p < .05) as the variables bearing a statistically significant influence over the number of innovations. However, the distance related variables returned no statistically significant influence on innovation incidence. Model 3 displays how firm age (B=-0.02, Z=-2.99, p < .01), manager age (B=0.02, Z=3.39, p < .01) and annual firm turnover in excess of €400,000 (B=0.23, Z=2.15, p < .05) all return a statistically significant influence on the number of innovations. In terms of the firm location characteristics, Centre region firms (B=0.23, Z=2.02, p < .05) attain higher numbers of innovations and proximity to Oporto boosts the total number of innovations according to a quadratic effect (B=0.14, Z=2.06, p < .05). The Vuong test reveals that Models 1 and 2 are equally statistically explicative, (p > .05) as are Models 1 and 3, however Model 3 is more statistically explanative than Model 2 at p < .10.

16

Table 5 – Count models: Dependent variable – Total number of innovations Model 1

Model 2

Model 3

B

SE

Z

B

SE

Z

B

SE

Z

Intercept

-0.13

0.23

-0.58

0.05

0.25

0.20

-0.22

0.25

-0.88

t_KIBS

-0.04

0.05

-0.66

-0.03

0.06

-0.48

-0.02

0.05

-0.43

Firm age

-0.02

0.01

-3.49***

-0.02

0.01

-3.32***

-0.02

0.01

-2.99***

Manager age

0.02

0.01

3.51***

0.02

0.01

3.45***

0.02

0.01

3.39***

TO > €400,000

0.24

0.10

2.33**

0.25

0.11

2.40**

0.23

0.11

2.15**

No. employees

0.00

0.00

-0.81

0.00

0.00

-1.20

0.00

0.00

-0.79

Graduate employee (%)

0.04

0.11

0.39

0.02

0.11

0.17

0.06

0.10

0.54

North

0.20

0.07

2.88***

0.43

0.34

0.18

Lisbon

0.21

0.09

2.26**

0.35

0.19

0.21

Centre

0.14

0.08

1.79*

0.23

0.11

2.02**

Algarve

0.19

0.10

1.89*

0.20

0.17

0.12

log10(DL)

-0.05

0.11

-0.48

-0.12

0.26

-0.45

log10(DP)

-0.07

0.07

-0.99

0.12

0.11

1.06

2

-0.01

0.08

-0.19

-0.07

0.11

-0.70

2

0.05

0.05

1.08

0.14

0.07

2.06**

(log10(DL)) (log10(DP)) n Pseudo-R2 - 2 LL

437 0.077 33.01

AIC BIC

1289.0 1333.9

V test (M1 vs M2)

0.37

V test (M1 vs M3)

-1.12

V test (M2 vs M3)

-1.33*

437 0.074

437 0.085

31.79 1290.2 1335.1

36.73 1293.3 1354.5

Note: * p < .10, ** p < .05, *** p < .01; B: Regression coefficient; SE: Standard error; Z: Z-test; TO > €400,000 – Turnover greater than €400,000; Graduate employees (%): Percentage of employees 2

holding university level education qualifications; DL: Distance to Lisbon; DP: Distance to Oporto; Pseudo-R : Cragg and Uhler's pseudo r-squared; -2LL: -2 Log likelihood; AIC: Akaike Information

Criterion; BIC: Bayesian Information Criterion; V Test: Vuong test

17

Final Considerations

The literature review broadly demonstrated the scope of the role undertaken by KIBS firms and how their presence and activities are important both to innovation taking place at other firms and innovation in the host region. This research project correspondingly sought to address the effects of location on KIBS firms and their innovative capacities. We thus sought to verify whether firm locations closer to the major Portuguese urban centres (Lisbon and Oporto) experience a positive effect on their innovative capacities. The empirical evidence reports that in the case of service innovation, the fact of firm location being further from these main markets negatively influences both this innovation type as well as the total number of innovations. Furthermore, the older the manager/owner positively influences, that is raises the incidence, of this innovation type and a factor that also remains valid for process innovations and for the total number of innovations processed by the firm. As regards the KIBS location in a specific region, we conclude that firms located in the North, Lisbon and Centre regions innovate their services with a greater degree of frequency than firms located in the Algarve and Alentejo. As regards the Centre region, we should highlight that firms located here, when analysing the total number of innovations, innovated more than in all other Portuguese regions. As regards the distance from the two main urban centres (Lisbon and Oporto), in the case of Lisbon, there were no statistically significant influences registered while in the case of Oporto, our findings show that the greater the KIBS proximity to this urban centre, the greater the innovation in services, in processes as well as in the total number of innovations. We should highlight that we do not report statistically significant evidence as regards innovation by KIBS type, professional or technological. These conclusions corroborate some of the findings in the literature in demonstrating that the closer the firm is located to the major urban centres, the greater the level of their innovations (Markusen, 1999; Keeble & Nachum, 2002; Cooke et al., 2004; Doloreux, 2004). As regards financial performance, we also conclude that this interrelates with the number of innovations. Hence, locations near urban centres not only foster innovation but also serve to boost firm financial performance and a finding in harmony with some authors in the literature (Kostopoulos et al., 2011; Forsman, 2011). 18

This study, we believe, contributes towards a better understanding of the innovative behaviours of KIBS firms and the effects caused by their location. Given entrepreneurship and innovation prove fundamental to regional development, this raises the need for the design of support and incentive policies for the founding and running of firms away from the major urban areas. Indeed, national development will only ever be able to attain the equal progress of the diverse different regions through the provision of effective and efficient innovation systems (Porter & Steiner, 2001). This study does find that the most innovative KIBS firms are those located within the vicinity of the main urban centres. Ascertaining the effect of location on KIBS firm innovation capacities in different international geographic contexts thereby becomes, in our perspective, a clear priority for future research.

Bibliography Aslesen, H., Isaksen, A. (2007a) Knowledge intensive business services and urban industrial development, Service Industries Journal, 27(3): 321–338. Aslesen, H., Isaksen, A. (2007b) New perspectives on knowledge-intensive services and innovation, Geografiska Annaler, Series B: Human Geography, 89B (Suppl. 1): 45–58.

Aslesen, H., Jakobsen, S. (2007) The role of proximity and knowledge interaction between head offices and KIBS, Tijdschrift voor Economische en Sociale Geografie, 98(2):188–201. Becker, W., Dietz, J. (2004) R&D Cooperation and Innovation Activities of Firms – Evidence for the German IPO Data, Economics of Innovation and New Technology, 15(1): 71-81. Camacho, J. A., Rodriguez, M. (2005). How innovative are services? An empirical analysis for Spain, Service Industries Journal, 25(2): 253–271. Chadwick, A., Glasson, J. (2008) Employment growth in knowledge-intensive business services in Great Britain during the 1990s—variations at the regional and subregional level, Local Economy, 23(1): 6–18. Coe, N. (1998) Exploring uneven development in producer service sectors: Detailed evidence from the computer service industry in Britain, Environment & Planning A 30: 2041–68. Cohen, W.M., Levinthal, D.A. (1990) Absorptive capacity: a new perspective on learning and innovation. Administrative Science Quarterly 35 (1): 128–152. Cooke, P. (2001) Strategies for Regional Innovation Systems, Policy paper,Vienna, United Nations Industrial Development Organization (UNIDO). Cooke, P., Heidenreich, M., Braczyk, H-J (Eds) (2004) Regional Systems of Innovation, London: Routledge. Cooke, P., Leydesdorff, L. (2006) Regional development in the knowledge-based economy: The construction of advantage, Journal of Technology Transfer 31: 5–15. Coffey, W., Shearmur, R. (1997) The growth and location of high-order services in the Canadian urban system, 1971–1991, The Professional Geographer 49: 404-18. Czarnitzki, D., Spielkamp, A. (2003) Business Services in German: Bridges for innovation, The services industries Journal, 23 (2): 1-30. 19

Doloreux, D., Dionne, S. (2008) Is regional innovation system development possible in peripheral regions? Some evidence from the case of La Pocatière, Canada, Entrepreneurship & Regional Development, 20 (3): 259-283. 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 Ferreira, J. (2010) Corporate entrepreneurship and small firms growth, International Journal of Entrepreneurship and Small Business 10 (3): 386 – 409. Forsman, H., (2011) Innovation capacity and innovation development in small enterprises, A comparison between the manufacturing and service sectors, Research Policy 40 (5): 739–750. Fosfuri A., Tribó, J.A. (2008) Exploring the antecedents of potential absorptive capacity and its impact on innovation performance, Omega, 36: 173–87. Freel, M. (2006) Patterns of Technological Innovation in Knowledge-Intensive Business Services, Industry and Innovation, 13(3): 335-358. Gallouj, F., Weinstein, O. (1997) Innovation in services, Research Policy, 26: 537-556. Gong, H. (2001) A hierarchical change model of business and professional services in the United States, Urban Geography 22: 340–59. He, Z., Wong, P. (2004) Exploration vs. Exploitation: An Empirical Test of the Ambidexterity Hypothesis, Organization Science, 15 (4): 481-494. Howells, J.R.L., Tether, B.S. (2004) Innovation in Services: Issues at Stake and Trends. Bussels, Luxembourg: European Commission. Hu, T., Chang, S., Lin, C., Chien, H. (2006) Evolution of knowledge intensive services in a high-tech region: the case of Hsinchu, Taiwan, European Planning Studies, 14(10): 1363–1385. Huergo, E., Jaumendreu, J. (2004) How does probability of innovation change with firm age? Small Business Economics, 22(3-4): 193-207. Hull, C.E., Covin, J.G. (2010) Learning capability, technological parity, and innovation mode use, Journal of Product Innovation Management 27 (1): 97–114. Hult, G.T., Hurley, R.F., Knight, G.A. (2004) Innovativeness: its antecedents and impact on business performance, Industrial Marketing Management 33(5): 429– 438. Johansson, B., Karlsson, C., Stough, R. (2001) Theories of Endogenous Regional Growth, Lessons for Regional Policies, Berlim; Springer – Verlag. Keeble, D., Nachum, L. (2002) Why do business service firms cluster? Small consultancies, clustering and decentralization in London and southern England, Transactions of the Institute of British Geographers NS27:67–90. Kline, S.J. (1985) Innovation is not a linear process, Research Management 28 (4): 36– 45. Koch, A., Stahlecker, T. (2006) Regional Innovation Systems and Foundtion of Knowledge Intensive Business Services, European Planning Studies 14(2): 123146. Kostopoulos, K., Papalexandros, A., Papachroni, M., Ioannou, G. (2011) Absorptive capacity, innovation, and financial performance, Journal of Business Research 64: 1335-1343. Lafuente, E., Vaillant, Y., Serarols, C. (2010) Location decisions of knowledge-based entrepreneurs: Why some Catalan KISAs choose to be rural? Technovation 30: 590–600.

20

Lynskey, M.J. (2004) Determinants of innovative activity in Japanese technology-based start-up firms, International Small Business Journal 22 (2): 159-196. Miles, I. Kastrinon, N. Flanagan, K. Bilderbeek, R. den Hertog, P., Huntink, W., Bouman, M. (1995) Knowledge intensive Business services, Users and sources of Innovation, Brussels: European Comission. Mitchell, R. K., Busenitz, L., Lant, T., McDougall, P. P., Morse, E. A., Smith, J. B. (2002) Toward a theory of entrepreneurial cognition: rethinking the people side of entrepreneurship research, Entrepreneurship Theory & Practice, 27(2): 93-104. Moyart, L. (2005) The role of producer services in regional development: What opportunities for medium-sized cities in Belgium? Services Industry Journal 25: 213–28. Malecki, E., Nijkamp, P., R. Stough, R. (2004) Entrepreneurship and space in the network age (Special issue editorial), Entrepreneurship and Regional Development 16: 1–3. Muller, E. (2001) Innovation Interactions Between knowledge intensive business and small and medium-sized enterprises, Heidelberg, New York: Physica- Velarg. Muller, E., Doloreux, D. (2009) What we should know about knowledge-intensive business services, Technology in Society 31: 64–72. Nelson, R., Winter, S. (1982) An Evolutionary Theory of Economic Change, Belknap Press of Harvard University Press, Cambridge (MA). Nieto, M. J., Santamaria, L. (2005) Novelty of product innovation: the role of different networks. Business Economics Series, Universidad Carlos III de Madrid. Working Papers #05-65. November. Nonaka, I., Takeuchi, H. (1995) The Knowledge Creating Company. How Japanese Companies Create the Dynamics of Innovation, Oxford University Press, New York. OhUallach´ain, B., Reid, N. (1991) The location and growth of business and professional service in American metropolitan areas, 1976–1986, Annals of the Association of American Geographers 81: 254–70. Pavitt, K. (1984) Sectoral patterns of technological change: towards a taxonomy and a theory, Research Policy 13: 343-373. Poehling, R. (1999) Locating producer services in the rural South of the US: The case of Alabama, Florida and Georgia. Paper presented at the Meetings of the North American Regional Science Association, Montreal, Canada. Román, J., Gamero, J., Tamayo, J. (2011) Analysis of innovation in SME using na innovative capability-based non-linear model: A study in the province of Seville (Spain). Technovation, 31: 459-475. Romer, P.M. (1994) The Origins of Endogenous Growth. Journal of Economic Perspectives, 8(1): 2–22. Shearmur, R., Doloreux, D. (2008) Urban Hierarchy or Local Buzz? High-Order Producer Service and (or) Knowledge-Intensive Business Service Location in Canada, 1991-2001, The Professional Geographer, 60(3): 333-355. Shearmur, R. (2011) Innovation, Regions and Proximity from neo-regionalism to spatial analysis, Regional Studies, 49: 1225-1243. Simmie, J., Strambach, S. (2006) The contribution of KIBS to innovation in cities: And evolutionary and institutional perspective, Journal of Knowledge Management 10: 26–40. Sundbo, J. (1998) The Organization of Innovation in Services (Frederiksberg, Denmark: Roskilde University Press).

21

Teece, D.J., Pisano, G., Shuen, A. (1997) Dynamic capabilities and strategic management, Strategic Management Journal 18 (7), 509–533. Tether, B., Hipp, C., Miles, L. (2001) Standardisation and particularisation in services: evidence from Germany, Research Policy, 30: 1115-1138. Tether B., Hipp, C. (2002) Knowledge intensive technical and other services: patterns of competitiveness and innovation compared, Technol Anal Strategic Manag 14(2): 163–182. Tether, B. S. (2003) The Sources and Aims of Innovation in services: variety between and within sectors, Economics of Innovation and new technology 12(6): 481-505. Toivonen, M. (2004) Expertise as Business: Long- Term Development and Future Prospectives of Knowledge Intensive Business Services (KIBS). Doctoral Dissertation Series 2004/2, Helsinki University of Technology, Department of Industrial Engineering and Mannagement. Wakelin, K. (1998) Innovation and export behaviour at the firm level, Research Policy 26 (7–8): 829–841. Wernerheim,M., Sharpe, C. (2003) “High-order” producer services in metropolitan Canada: How footloose are they? Regional Studies 37: 469–90. Wood, P. (2005) A Service – Informed Approach to regional innovation – or adaptation? The services Industries Journal 25( 4): 429-445. Yam, R.C.M., Lo, W., Tang, E.P.Y., Lau, A.K.W. (2011) Analysis of sources of innovation, technological innovation capabilities, and performance: an empirical study of Hong Kong manufacturing industries, Research Policy 40 (3): 391–402

22