The effect of human capital diversity on firm innovativeness within a knowledge intensive service industry

1

The effect of human capital diversity

on firm innovativeness within a

knowledge intensive service industry

Master Thesis

by

Peter van Spaandonk s1948059

Supervisors Florian Noseleit Thijs Broekhuizen

University of Groningen

Faculty of Economics and Business

2

Abstract

Literature on the effects of human capital diversity on the innovativeness of service organizations is scarce. This study aims at minimizing this literature gap alongside a

quantitative research approach. The Dutch marketing and communication sector serves as the research area. The effects of diversity in terms of age, gender, nationality and educational background among the total workforce on the innovative performance of the firms are tested by means of a questionnaire. Furthermore, the effects of the share of full-time employees and the share of highly educated employees have on the relationship between human capital diversity and firm innovativeness are examined.

The outcomes show that the surface-level diversity (a combination of diversity in age, gender and nationalities) contributes to the overall innovation process. Diversity in educational background does not directly support the innovative output of the organization as it supports service innovation but counteracts process innovation. The effects of surface-level diversity and educational background diversity on the organizations innovativeness are altered when moderated with the share of full-time employees and the share of highly educated employees. The results of this research imply that human capital diversity can positively influence a firm’s innovative performance but only when specific conditions are met.

Keywords: human capital diversity, age, gender, nationality, educational background, full-time

3

Pre-face

The finalization of my study in the form of this thesis was an highly satisfying but also

challenging period for me. I was very grateful with the chance to investigate a topic of my own interest and I enjoyed the freedom to design and conduct my study according to my own beliefs and thoughts. Matter of course I also felt the pressure to deliver an attractive thesis that thoroughly investigates the topic on hand and exposes relationships that were unclear in literature so far. On forehand I had high demands of myself. These high demands sometimes counteracted me a bit as my enthusiasm and expectations sometimes exceeded the real possibilities. A logical consequence of this is the extension of the intended amount of time to finalize my thesis. However, in the end I feel that overcoming these obstacles was part of a natural process and necessary to deliver a qualitative thesis.

I want to make use of this opportunity to thank my supervisor Florian Noseleit who has supported me in writing my thesis and provided helpful and guiding feedback. The meetings to discuss my documents and progress always gave me new insights and triggered my creativity. The informal and open attitude of the supervision in general was very supportive for my overall progress.

Moreover, I want to thank my family and friends for the support along the way. The

enthusiasm of my fellow students and the interest in my progress of my family boosted me in times when my progress slowed down alongside my optimism. In special I want to thank my friends and co-students with whom I spend long hours, days, weeks in the library. We supported and motivated each other along the way which in my opinion not only helped to finalize our theses but also significantly improved the quality of the work delivered.

4

Management summary

This study investigates the effects of human capital diversity on firm innovativeness within the Dutch marketing and communication industry. Human capital diversity in the realm of this research is determined by variation among the total workforce in terms of age, gender, nationalities and educational background. Moreover, the potential influence of the share of full-time employees and the share of highly educated employees have on the relationship between human capital diversity and firm innovativeness are tested.

The literature review shows that literature on the effects of diversity on organizational performance and innovation has not reached consensus on the nature of the effects. Some studies show positive effects of diversity on innovation while others claim that diversity may counteract the innovation process. This double edged sword generally emerges for each type of diversity entailed in this study, being: age diversity, gender diversity, nationality diversity and educational background diversity. Research on the effects of full-time workers on the innovativeness of the firm generally state that full-time workers are less productive as they spend more time on tasks not directly related to the firms productiveness. Studies on

educational degree in relation to innovative performance of teams show a positive relationship between the level of education and the innovation measures. However, a positive group climate is necessary to benefit from the supportive nature of highly educated individuals within a team.

The marketing and communication industry can be described as a knowledge intensive service industry meaning that knowledge and ability are the key aspects instead of clearly defined services. The main characteristic of this type of industry in the realm of innovation is that products are not always perfectly formatted and codified which makes innovation

5 product-services but competencies and abilities. The interface with the customer is an

important aspect within this innovation pattern.

Academic literature on innovation typology is extensive and diverse. In the realm of this study a separation between service innovation, process innovation and organizational innovation is made. Service innovation simply refers to the development of new services while process innovation is mainly attributed to the realizing greater efficiency. Organizational innovation includes new introductions or substantial alteration in business structures, strategy and management.

Besides innovation typology on basis of the type of innovation this study also differentiates between innovationson basis of the degree of newness embedded in the innovation.

To determine the effects of human capital diversity on innovation this study makes use of a questionnaire which is diffused among Dutch marketing and communication agencies. By means of this questionnaire, both the diversity among the workforce as well as the innovative performance are measured. Thereby the effects of age diversity, gender diversity and

nationality diversity on firm innovativeness are merged to determine the effect of human capital diversity. Furthermore the potential influence of the share of full-time employees and the share of high educated employees is investigated through this questionnaire.

On basis of data reduction techniques this study finally measures innovation alongside four dependent variables. Service innovation and process innovation are separately measured. Moreover, these two types of innovation are also combined into a new variable. Finally, innovation comprising a high degree of newness is included in the fourth and final dependent variable.

The results show that surface-level diversity, the aggregated diversity measure, is in general beneficial for each type of innovation included in the research. When a larger share of the workforce is employed on a full-time basis the effect of surface-level diversity on the overall innovation process, service and process innovation combined, turns negative. Diversity in educational background is supportive for service innovation but counteracts process

6 innovation. Innovation comprising a high degree of newness is positively influenced by a larger share of high educated personnel.

7

Index

1.1 A shift in the composition of the Dutch labour force ... 11

2. Problem statement ... 14

3. Theoretical reflection ... 15

3.1 Diversity: a diverse term ... 15

3.2 Measuring (maximum)diveristy ... 16

3.3 Surface-level diversity in relation to innovation ... 17

3.3.1 Age diversity ... 17

3.3.2 Gender diversity ... 19

3.3.3 Nationality diversity ... 20

3.4 Educational diversity ... 20

3.5 The effects of full-time and part-time employment on innovation ... 22

3.6 Innovation typology ... 23

3.7 Diversity and the radicalness of innovation ... 26

3.8 Diversity and innovation, a double-edged sword ... 26

3.9 Measuring firm’s innovativeness ... 28

3.10 Innovation in service industries ... 29

4. Research question and conceptual model ... 34

4.1 Hypotheses formulation ... 35

5. Methodology... 38

5.1 Research design ... 38

5.2 The questionnaire ... 38

8

5.3 Sample and data collection ... 39

5.4 Research approach method ... 40

6. Data and Measurement ... 41

6.1 Independent variables: Diversity measures ... 41

6.2 Moderating variables ... 42

6.3 Dependent variables: Innovation measures ... 43

6.3.1 Factor analysis ... 43

6.3.2 The Innovation Measures ... 45

7. Results ... 46

7.1 Summary of findings Table 1. Multiple Regression, No moderators ... 47

7.2 Summary of findings table 2. Multiple regression, moderator: % Full-time employees ... 48

7.3 Summary of findings table 3. Multiple regression, moderator: % High Educated employees ... 50

7.4 Summary of findings table 4. Multiple regression, moderators: % of Full-time employees and % of High Educated employees ... 52

7.5 Hypotheses testing ... 53

8. Conclusion and recommendations ... 55

8.1 Discussion of results ... 55

8.1.1 Surface-level diversity and Innovation ... 56

8.1.2 Educational Discipline diversity and Innovation ... 59

8.1.3 The moderators ... 63

8.2 Brief summary of most important findings and general conclusion ... 66

8.3 Managerial implications ... 68

8.4 Limitations and future research ... 70

References ... 72

Appendices... 81

Appendix A1: The questionnaire ... 81

9

Appendix B: Factor analysis outcome ... 87

Appendix C1: interaction effect model 1.1 ... 88

Appendix C2: interaction effects model 2.1 ... 89

Appendix C3: interaction effect model 3.2 & 4.2 ... 90

10

1. Introduction

Considering that workforce diversity has dramatically increased (Ragins and Gonzalez, 2003), practitioners acknowledge that having a diverse workforce may be a key for sustained competitive advantage by increasing creativity and innovation (Bassett-Jones, 2005; Richard, 2000). Looking at forecasts from the Dutch Central Bureau of Statistic (CBS) the Dutch labour force will change tremendously in the forthcoming years. Gordon and McCann (2005) state that “differences in the geography of creativity and entrepreneurship” are based on a diversity of skills, ideas and cultures that enable new combinations of knowledge to emerge, a permissive environment that allows different and unorthodox ideas to emerge and a highly competitive environment, including discriminating consumers of new products. Literature on labour market diversity in relation to innovation suggest that diversity in the labour market in which the business is located is associated with greater creativity and innovation at the

regional level (Grossman &Helpman, 1994; Florida, 2002; McCann &Simonen, 2005). Bearing in mind that a diversely composed workforce can be key for innovativeness and competitive advantage, this rapidly changing labour force offers an interesting and relevant research opportunity.

This proposed research will focus on the effect that diversity among the workforce of Dutch SME’s in the marketing and communication segment has on the innovativeness of the companies. In prior research several factors have been investigated in terms of their effect on a company’s innovativeness. It is self-evident that the workforce of a company for a great deal determines its innovative capacity. This makes it interesting to investigate whether the

composition of the workforce influences the innovativeness of a company. At first the forthcoming change within the Dutch labour force will be discussed in order to get a clear perspective on the immensity of this phenomenon. Subsequently the problem statement derived from this phenomenon will be given. The theoretical reflection covers academic literature on the relevant topics for this study followed by the central research question for this study together with the hypotheses and the conceptual model. Thereafter the

11

1.1

A shift in the composition of the Dutch labour force

The composition of the Dutch labour force is changing rapidly. This change is an interesting phenomenon as it gives employers the opportunity to diversely compose their workforce. Since this study focuses on the effects of human capital diversity on innovation the change in the composition of the Dutch labour force is an interesting aspect to highlight in the realm of this research.

The change in the composition of the Dutch labour force is reflected in several areas. Overall ageing, a tremendous growth in ethnic minorities and a trend of more highly educated people ensure a different composition of the labour force in the Netherlands for the forthcoming years. To get a clear perspective on this phenomenon, the next paragraphs will present figures and trends in the Dutch labour force referred to the variables to be measured in this study: age, gender, nationality and educational background.

The variable age is an interesting variable to take into account as the Dutch population, and therefore the Dutch labour force, is ageing significantly in the coming decades. The graph presented in figure 1 shows the aging process of the Dutch labour force. The percentage of workers aged 50-64 increases onwards from the early nineties. With labour force participation levels of 2000 kept constant, the percentage of older workers in the labour force will increase substantially in the following decades. Midterm projections of the Dutch Bureau for

Economic Policy Analysis show increasing participation rates of older males and particularly older females, the aging process will be much stronger and rapid (Remery et all, 2003).

Figure 1: Age distribution Dutch labour force, 2000-2050

12

Sex / Year 2010 2015 2025

Man 5604.000 5577.000 5467.000

Female 5519.000 5512.000 5438.000

Table 1: Dutch labour (aged 15-64) force divided by sex, 2010-2025

A deeper insight into the Dutch labour force in terms of nationalities of the workforce shows that the ethnic minority grows rapidly. This is mostly due to the fact that the ethnic minority in the Netherlands is in general younger compared to the native population. Nowadays twelve per cent of the labour force in the Netherlands consists of non-western foreigners and 10 per cent western foreigners. In 2050 these percentages have increased to twenty-one per cent non-western foreigners and 13 per cent non-western foreigners as a part of the Dutch labour force (Dutch Central Bureau for Statistics, 2011).

The diversity in educational backgrounds of employees of SME’s in the Netherlands can be determined in terms of educational level and educational disciplines. The average educational level among the Dutch labour force will increase in the coming decades. This is mostly due to fact that the older and mostly lower educated employees will be replaced by younger

employees with a higher level of education. Besides this cohort effect, the qualification requirement (the extension of compulsory education to the age of eighteen for people without basic qualifications) will also ensure that the degree of the population with only primary school will decrease. According to forecasts by the Dutch Social Economic Council the percentage of highly educated among the potential Dutch labour force will increase to thirty-six per cent in the year 2020, while the percentage of the population with only primary education will drop to six per cent in 2020.

The on-going trend of accomplishing a higher educational level among Dutch citizens ensures that the diversity in educational disciplines will also increase due to the broad range of

13 2005 2010 2020 Growth rate Primary education 10 9 6 -3,0 Lower education 22 20 18 -1,4 Medium education 40 40 40 0,0 Higher education 28 31 36 1,7 H/L-ratio 39 45 57 2,6

Table 2: Percentage distribution qualification structure of potential Dutch labour force, 2005-2020

The significant changes in the composition of the Dutch labour market offer possibilities for Dutch companies to improve their innovative output as innovation involves the creation of new knowledge arising from the interaction of different peoples, talents, interests, insights and experiences (Lundvall, 2009). Jacobs (1969) states that diversity is of great importance as it a source of knowledge for innovation, typically seen in urban areas, where this diversity facilitates the blending of different types of knowledge for what Schumpeter (1934) refers to as “new combinations”. This implies that organizations should face local labour market diversity as a source of knowledge for innovation and policy-makers should likewise embrace diversity as a support for business innovation (McGuirk and Jordan, 2012)

14

2. Problem statement

The Dutch labour force faces a tremendous ageing process in the forthcoming years. Besides the ageing process the ethnic minorities grow rapidly alongside a shift in the average education level of the labour force (www.cbs.nl, 2011). As stated before, several practitioners

acknowledge that having a diverse workforce may be key for sustained competitive advantage by increasing creativity and innovation. This offers an interesting research possibility as companies within the Netherlands have the possibility to compose and to utilize such a diverse workforce. As the marketing and communication is a business segment in which innovativeness is crucial for success this study focuses on this specific segment. The goal of this study is to determine whether the diversity in the composition of the workforce influences the innovativeness of marketing and communication orientated SME’s in the Netherlands. Diversity in the realm of this study focuses on the previously discussed

determinants age, gender, nationality and educational background. Furthermore, the potential influence of part-time employment and the level of education attained will be examined. Derived from the problem statement is the central research question:

What is the effect of human capital diversity on the innovativeness of Dutch marketing and communication agencies?

15

3. Theoretical reflection

This section covers the theoretical reflection of diversity among groups and total workforces in relation to innovativeness of companies. Much has been written in academic literature on the relation between for example diversity in top management teams and work groups and the innovativeness of these groups or companies. The effect of diversity of the total workforce on the innovativeness of the company has been researched to a lesser extent. Since diversity is a general term the first part of this theoretical reflection will further specify the meaning of diversity for this study. Thereafter the measurement of diversity will shortly be discussed. As the relation between age diversity, gender diversity, nationality diversity and educational diversity and the innovative outcome of a group and also companies has been widely discussed with some interesting, sometimes contradicting findings, the theories on this relationships will be reflected separately. Subsequently, theory on the possible influence of part-time versus full-time employment on organizations performance and innovativeness will be treated. Thereafter, a comprehensive innovation typology is described in the following section to be followed by a brief description of the relationship between diversity and more radical innovation and a description of the “double-edged sword” that emerges in literature on diversity and innovation. Subsequently the measurement of firm innovativeness is shortly discussed and the literature review is finalized by a description of the complexity of innovation within service industries.

3.1

Diversity: a diverse term

Organizations continue to be more diverse in the sociological and psychological characteristics of their workforce (Nemetz and Christensen, 1996; Offerman and Gowing, 1990). There continues to be a strong need to understand the impact of diversity on individual and team processes and outcomes in the workplace. The number of articles published has grown almost exponentially in the past two decades. This trend shows that diversity is commanding a great deal of research attention ( Harrison and Sin, 2006).

16 adopts an individual focus by referring to diversity as the perceived differences from the self (Roberge and van Dick, 2010).

All though the fact that diversity is an increasingly vital and pervasive topic, there is no existence of a favored or dominant (set of) approach(es). Likewise, there is no constitutive definition for the term diversity (Alison et al., 2006) Alison et al. (2006) in their Handbook on Workplace Diversity deal with both of the impediments. In general they state that the best approach greatly depends on scale level of the variable studied. Each scale level ask for a different measurement method, this will be further specified in the next section of this literature review. As for the definition of the term the authors define diversity as: a collective amount of differences among members within a social unit. Furthermore, the authors state that diversity is meaningful only when it is more narrowly defined or dimensionalized. In order for the term to be a viable construct in research it needs to be accompanied by some adjective or modifier e.g. sex diversity, age diversity, cognitive ability diversity, educational background diversity and so on. In terms of dimensionalizing diversity, more abstract constructs have already began to percolate through the literature.

A group can be considered diverse when it is constituted of individuals that belong to and identity themselves with different subgroups. People may be different from one another based on a large number of indicators of diversity. The existence of a diverse workgroup is not only multidimensional but may signify many implications (Roberge and van Dick, 2010). In order to get a more thorough understanding of the different indicators of diversity in groups Harison, Price and Bell (1998) proposed two distinct dimensions of diversity indicators: surface-level diversity and deep-level diversity. The authors have defined surface-level diversity as “difference s among group members in overt, biological characteristics that are typically reflected in physical features”. Surface-level diversity refers to the attributes that are visible and easily perceived by individuals such as age, gender and race/ethnicity. Other visible stigma such as physical handicap, disfigurement, as well as weight problems (i.e. obesity or anorexia) could also be classified under surface-level diversity. As opposed to surface-level diversity deep-level diversity refers to more subtle attributes that cannot necessarily be directly and immediately observed. Such attributes refer to personality, attitudes, beliefs and values but also sexual orientation and religious beliefs.

3.2

Measuring (maximum)diversity

17 Workplace Diversity demarcate three possible sets of beliefs regarding what a maximum collective amount of differences entails. Maximum diversity could mean that one person reigns over all the rest of the members in a social unit, creating a strongly positively skewed distribution. It could also mean that a social unit contains subgroups occupying opposite endpoints of some continuum. A third meaning of maximum diversity might be that each member within the social unit takes a different category value, or is equally spaced from each of the contiguous members. Alison et al. (2006) describe this type of diversity as “variety”. Each meaning asks for a different measurement approach. As this study measures diversity alongside categorical values the Blau Diversity Index is used, the most commonly employed measure for diversity as variety (e.g. Klein et al., 2001; Timmerman, 2000; Bunderson & Sutcliffe, 2002). The underlying computations will further be specified in the methodology section.

3.3

Surface-level diversity in relation to innovation

Based on the previous section clarifying diversity for the realm of this study this section separately deals with literature on the surface-level diversity aspects age, gender and nationality in relation to innovation.

3.3.1 Age diversity

An important notice beforehand is to mention the intertwining between innovative

performance and overall performance in the existing literature on age related differences on innovativeness. Age related differences in productivity can either result from age-specific work performance in general, or emerge from an age-specific capacity to develop or adopt new, productivity-enhancing technology. As a consequence, these two effects cannot be properly disentangled (Frosch, 2011).

Management literature extensively covers the interplay between age diversity in teams and performance (Horwitz and Horwitz, 2007; Riordan, 2000; Wegge and Schmidt, 2009; Williams and O’Reilly, 1998), but rarely focuses on innovation as an outcome. Until now empirical evidence onwhether age diversity is conducive to team performance is inconclusive (Frosch, 2011). Bantel and Jackson (1989) clarify the problem: different age cohorts of workers have different attitudes and values. On the one hand, diversity in teams can serve as the backdrop for team performance. The similarity – attraction paradigm and social identity theory thereby posits that workers communicate more frequently and effectively if they are similar in age and tenure, as they share more common attitudes. On the other hand, if diversity fosters team creativity, groups composed of member of various ages should be more innovative.

18 and Schmidt, 2009; Wegge et al., 2008) and those requiring creative behaviour (Choi, 2007; Thornburg, 1991; Triandis et al., 1965). However, with respect to the age diversity in top management teams, Wiersema and Bantel (1992) did not find effects on corporate strategic change in a sample of Fortune 500 companies, and Bantel and Jackson (1989) fail to find effects for innovation in banking.

The studies mentioned above either focus on general creative behaviour or focus on age diversity in top management teams. This proposed study is focussing on the diversity of the total workforce of marketing and communication bureaus in the Netherlands. Therefore it is interesting to look at empirical findings on the linkage between the age diversity in the overall workforce and innovation. Several studies have indeed focused on the age diversity among the overall workforce of a firm and its performance in industrial innovation. The existing studies on the productivity effects of age heterogeneity in the workforces of firms identify either a hump shaped relationship between productivity and age diversity(Grund and Westergard-Nielsen, 2008) or no relation at all (Ilmakunnas et al, 2004). One study has focused explicitly on the effect of age diversity on innovative performance. This study by Veen and Backes-Gellner is based on employer-employee data set referring to several thousand German firms between 1993 and 2001. Veen and Backes-Gellner conclude that the more age diverse a firm’s workforce is, as measured by two alternative diversity indices, the higher its productivity. According to their analysis, heterogeneity in the age composition of the workforce that is 10 percentage points higher would ceteris paribus lead to an increase in productivity of

approximately 3 per cent. In contract to this result, age-heterogeneity is found to be

detrimental for productivity in more traditional sectors. The authors explain their findings by speculating that inefficiencies in cooperation and communication between employees may increase with the degree of age heterogeneity in the workforce. However, if standardized routine tasks dominate, innovative firms seem to overcompensate for these influences

through improved and flexible problem-solving in a dynamic context, which may be attributed to a higher diversity of cognitive skills and other competences provided by mixed-age

workforces (Frosch, 2011). The results are justified in a robust manner against the inclusion of firm-fixed effects that cope with the potential omission of unobserved, time-invariant

19 3.3.2 Gender diversity

The effect of gender diversity on team processes and performance has been subject to a numerous empirical studies, meta-analyses and literature reviews (e.g. Baugh and Graen 1997; Bowers et al. 2000; Chatman and O.Reilly 2004; Ely and Thomas 2001; Jackson et al. 2003; Joshi and Roh 2009; Mannix and Neale 2005; Myaskovsky et al. 2005; Pelled 1996; Stewart 2006; Webber and Donahue 2001). In a study by Milliken and Martins (1996) on the

relationship between gender diversity and group cognitive functioning such as innovation they found no significant relation. In contrast an earlier study by Hoffman and Maier’s (1961) showed that gender diversity promoted innovation within teams. More recently, Rogelberg and Rumery (1996) showed that the number of males on a team positively influences the decision making quality. Furthermore they found that teams with a lone female outperformed all-male teams, suggesting that gender diversity adds to quality.

20 3.3.3 Nationality diversity

A limited amount of research is dedicated to diversity in nationalities, race and ethnicity in relation to group performance or innovative capacity. In the research that has been conducted in general a positive effect of racial diversity on innovation is exposed. Miliken and Martins (1996) found that racial diversity may be positively associated with some group-level cognitive outcomes such as the quality of ideas generated. Two studies conducted by McLeod et al. (1996) and Watson et al. (1993) offered evidence of diversity benefits in terms of quality of the solutions presented or ideas generated by ethnically diverse groups. McLeod et al. (1996) found that ethnically heterogeneous teams produced higher quality ideas in a brainstorming task that did homogeneous teams, although they did not necessarily produces more ideas or a greater number of unique ideas.

In a study by McQuirk and Jordan (2012) diversity in nationalities among the labour force showed to have a positive influence on product innovation among Irish manufacturing businesses. In contrast, the same study reported that diversity in terms of nationalities

negatively influenced process innovation. Meaning that greater levels of diversity in nationality reduce the likelihood of an organization engaging in process innovation.

3.4

Educational diversity

Educational diversity in the realm of this study focuses on the diversity in educational disciplines, the higher the number of different educational disciplines within an organization the higher the educational diversity. The literature on functional diversity within teams suggests that highly diverse teams decrease development time by increasing goal congruence among the functional groups, bringing more creative potential to problem solving and ensuring the availability of crucial input (Kragozoglu and Brown, 1993; Sethi et al., 2001). A study by Zirger and Hartley (1996) showed that for each additional function included on a product development team 0,4 months was subtracted from development time. In contrary, functional diversity may also increase cycle time. Research suggests that highly diverse groups have difficulty in developing a shared purpose and an effective group process and hence, fall down on implementation. A negative relationship between functional diversity and adherence to schedules is reported by Ancona and Caldwel (1992). Studies by Kessler and Chakrabarti (1999) and Sarin and McDermott (2003) found a lack of association between functional diversity and innovation speed. In line with this finding Carbonell and Rodriguez (2005) demonstrate an inverted U-shaped relationship between functional diversity and innovation speed.

21 that diversity of functional knowledge contributes positively to innovation. A study taking a broader scope is conducted by Ostergaard, Timmermanns and Kristinsson (2009). In their study on employee diversity’s impact on innovation they state that educational diversity has a positive effect on innovation with regard to diversity in educational background among the higher educated employees. Two types of dynamics influence the likelihood to innovate according to their study: different education types, or a more balance between education groups. A possible bias in their study regarding educational diversity lies in the fact that it measures diversity within the highly educated group. Employing higher educated employees would be positive for innovation performance while have more different types of educational backgrounds would increase the likelihood to innovate. Attained educational level is generally assumed to be correlated with cognitive ability and therefore higher levels of education should be associated with a team’s ability to generate (and implement) creative solutions to complex problems (Bantel and Jackson, 1989). This ability to generate creative solutions may explain why people who are more educated have more receptive attitudes towards innovation (Kimberley and Evanisko, 1981; Rogers and Shoemaker, 1971). However, higher cognitive ability among social members within a group does not always have to be beneficial for collaboration and thus innovative output. Badke-Schaub et al. (2010) in their study on cognitive conflict and the creation of creative ideas rely on the similarity-attraction paradigm by Byrne (1971) to explain the effect of cognitive ability on innovation. The similarity-attraction paradigm (Byrne, 1971) states that differences in attitudes and interests are

correlated with personal dislike and avoidance while similarity is linked with social attraction, meaning that people are attracted to those they perceive to be similar to them (Sing & How, 2000). The basic explanation for preference of similarity can be related to the concept of ‘cognitive dissonance’ (Festinger, 1957). Cognitive dissonance is the activated state of a person who experiences different or contradicting cognitions, beliefs, attitudes and behaviors. People strive to eliminate or reduce cognitive dissonance as it causes discomfort. Choosing similar team members seems to be a guarantee to avoid exposure to extremely different beliefs and attitudes. Empirical investigations on project teams point out that a positive group climate must be established to ensure information exchange among its members, enhancing creative insights.

22

3.5

The effects of full-time and part-time employment on innovation

Having discussed the relevant literature on human capital diversity this section covers theory on the effects of full-time and part-time employment on innovation. The effect of full-time and part-time employment on innovation is an interesting aspect in realm of this study as the Dutch marketing and communication industry relatively employs a high share of part-time employees. Most of these part-time employees are only deployed when customer demand peaks. This makes it an interesting aspect to investigate in the light of the effects on innovation within this industry.

Many researchers, mostly in the psychology area, have addressed the topic of full-time (FT) versus part-time (PT) employment. However, the literature comparing FT and PT employees mainly focusses on differences with regard to the personal characteristics of the employees, job satisfaction, job attitudes (Thorsteinson, 2003), turnover (Martin and Sinclair, 2007) and other variables related to the individual. Studies addressing the differences between FT and PT employees with regard to organizational performance or even individual performance are rare to be found.

An interesting FT versus PT employment research topic with regard to this study is the job attitude and job satisfaction aspect. This study aims at identifying the potential influence PT employment has on the relationship between workforce diversity and organizational

innovativeness. Job attitude is an important element in this examination as a poor job attitude directly influences creativity of the employee and will therefore decrease a company’s potential to be innovative. Literature on differences in job attitudes between FT and PT employees (FTE’s and PTE’s) show inconsistent results with some reporting that part-time workers have lower job satisfaction (Miller and Terborg 1979; Shockey and Mueller, 1994), some reporting higher job satisfaction (Barker, 1993; Eberhardt and Shani, 1984; Peters, Jackofsky, and Salter, 1981; Sinclair et al., 1999; Wotruba, 1990), and yet others finding no significant difference on job satisfaction (Krausz, Sagie, and Bidermann, 2000; Levanoni and Sales, 1990; Logan et al., 1973; Steffy and Jones, 1990).

Given that the literature did not reach a consensus on whether attitudinal differences exist between FTE’s and PTE’s , Thorsteinson (2003) attempts to quantify this relationship using meta-analysis. In particular, Thorsteinson (2003) examined possible differences between FT and PT workers on overall job satisfaction, facets of job satisfaction, organizational

commitment, job involvement and intention to leave. The results indicate that there are no overall differences between FT and PT employees with regard to job satisfaction,

23 employees are.

Nelen et al. (2009) analyze the effect of part-time employment on productivity in service sector firms using matched employer-employee data on Dutch pharmacies. Their research indicates that service firms with a high-share of part-time employment are more productive than firms with a high share of full-time employment. They relate these findings to the negative effect longer working hours have on the amount of time spent on demand-related tasks, the tasks directly related to customer demand. Their research shows that FTE’s spend relatively more time on overhead tasks, the tasks not directly related to customer demand such as administrative tasks and management tasks. They further discuss that the differences in job content between PTE’s and FTE’s might explain the productivity differential between PTE’s and FTE’s. Due to the flexibility that part-time jobs offer, part-time workers are expected to be deployed at times when customer demand peaks or is expected to peak. As a result, part-time workers are very likely to spend relatively more part-time on demand-related task than full-timers. These demand-related tasks can be considered as the most productive tasks as they correlate directly with the workload, unlike the overhead tasks. Therefore, the findings the study suggest that in the service-sector, part-time employment is beneficial for firm

productivity because part-time workers are deployed at times when customer demand peaks.

3.6

Innovation typology

Having described the different types of diversity and the effects of full-time and part-time employment on innovation this section discusses the innovation typology adopted for this study.

24 selectors” simplifying this definition into: “An innovation is something new with added value”. With this general definition of innovation Jacobs further describes the diversity that innovation contains by separating three forms of innovation; product (or service) innovation, process innovation and transaction innovation.

Product innovation simply refers to the development of new products and services and is considered the most important form of innovation by Jacobs. Without product innovations process and transaction innovations miss a purpose. Due to the variation in kinds of products there are also different kinds of product innovation, namely: new material products, new information products and new services. In the realm of this study only new service development has relevance. This form of innovation involves the introduction of a service that is new or substantially improved in terms of functional characteristics, technical abilities, ease of use, design or any other dimension.

Process innovations are changes in the production processes of products or services which in principle should lead to more efficient production. Eventually new or significantly improved production methods implemented as a result of process innovation reduce costs or increase the quality of the product or service delivered. Process innovation can be mainly technical where it relates to the actual production process for products or services but can also be of an organizational nature, where it involves the creation of business structures, practices and models. For the purpose of this study these forms of innovation are distinctively detached. Process innovation covers new or significantly improved production methods for services. Organizational innovation includes new introductions or substantial alteration in business structures, strategy and management. Despite its acknowledged importance, organizational innovation has generally received less attention in the literature than technological innovation (Sapprasert and Clausen, 2012). Only a few prominent studies consider the importance of organizational innovation (Fagerberg, 2004; Martin, 2008; Fagerberg and Sapprasert, 2010). One reason for this is the complexity of the relationship between organization and innovation and the absence of a single coherent theoretical framework for understanding the

25 measured in a consistent manner across industries and countries (Sapprasert and Clausen, 2012). To address these problems, this study adopts the definition and empirical

operationalization of organizational innovation in the Community Innovation Survey and the Oslo Manual (OECD, 2005), which have become established standard for innovation measurement. The third revision of the Oslo Manual (OECD, 2005) as implemented in the fourth Community Innovation Survey defines organizational innovation as “a new or

significantly improved knowledge management system implemented to better use or exchange information, knowledge, and skills within the firm” and/or “a major change to the

organization of work within the firm, such as change in the management structure or the integration of different departments or activities” and/or “a new or significant change in the firm’s relationships with other firms or public institutions, such as through alliances,

partnerships, outsourcing or sub-contracting”.

The last sub-category of innovation by Jacobs is transaction innovation and refers to new ways of both bringing products to the attention of consumers and selling them. Therefore innovations in this category take place in the realm of publicity, marketing and sales. In the context of this study transaction innovation is considered as a part of organizational

innovation as this already covers the overarching aspects structure, strategy and management. Besides classifying innovation on basis of the nature of the innovation a classification based upon the so called degree of newness is also very common in literature. Garcia and Calantone (2002) developed a clear and comprehensive typology for innovation based upon the degree of newness. The authors distinguish between marketing and technological innovation on either macro or micro level. Innovation on macro level are disruptive to their market, industry or even the whole world. Micro level innovations are only new to the existing resources of a company (Garcia and Calantone, 2002). Within these levels Garcia and Calantone (2002) state two perspectives in which the creation of discontinuities by the innovation is investigated. On one side the creation of discontinuities in the way of marketing a product or service. On the other side the discontinuities caused by the effect of a technological innovation. From these perspectives the authors construct three types of innovation; radical innovation, really new innovation and incremental innovation.

When further elaborating these perspectives a radical innovation can be classified as

26 environment. Finally, an incremental innovation refers to improvements to the existing technology in the existing market (Garcia and Calantone, 2002).

3.7

Diversity and the radicalness of innovation

As discussed in the previous section innovation has also been classified on basis of the degree of newness embedded in the innovation. This section briefly discusses the relationship between diversity and the radicalness of innovation.

Some researchers state that higher diversity leads to more radical innovation while others claim the opposite. March (1991) for example, while not thoroughly explaining the construct of diversity in his paper, claims that diversity among social members of a group leads to more explorative innovation while social homogenous groups tend to be more exploitative in nature. In contrast, some authors claim that higher levels of team diversity creates more conflict, counteracting the decision-making process that is of vital importance to achieve more radical innovation. Hoegl et al. (2003) in their study state that team diversity is beneficial in less innovative (less radical) projects in a more predictable working environment as tasks can be clearly assigned to each individual which reduces the possibilities for conflict among the different perspectives of the team. Nevertheless, in the case of more radical innovation, where the members of the team deal with a high degree of uncertainty, collaboration in the team is critical and the conflict associated with diversity can become an obstacle. In these cases, the benefits may even be surpassed by its disadvantages. Furthermore, Lovelace et al. (2001) state that when more diverse teams have excessive conflict and discussion without reaching consensus, the diversity hampers radical innovation.

The double-edged sword that emerges from literature on diversity and radical innovation is not unique for this topic butis generally reflected in innovation literature in relation to diversity. The following section briefly describes this.

3.8

Diversity and innovation, a double-edged sword

Prior research on information and decision making suggests that diversity among the workforce leads to improved decision-making processes by increasing creativity and

27 implementation of multidisciplinary teams, that is, cross-functional teams which combine representatives of different organization functions to ensure diversity in knowledge, perspective and mind-set (Roberge and van Dick, 2010)

However, studies providing evidence of increased group performance due to diversity among the group members are rare. Studies supporting such predictions have examined

informational-related attributes such as education, occupation or functional background (Jehn et al., 1999; Pelled et al., 1999; Wanous and Youtz, 1986). Amidst the studies that seek to understand demographic indicators, personality, values or attitudes only few have shown that diversity may indeed lead to positive outcomes (Bantel and Jackson, 1989; Barrick et al., 1998; Barry and Stewart; 1997).

In general the studies arguing that group diversity has a positive impact on innovativeness state that groups consisting of heterogeneous member entail more creativity. According to Nemeth (1986) heterogeneous groups incorporating minority members more carefully think about issues which results in better decision making and more creative ideas. The examination of the relationship between a culturally diverse workforce and organizational performance by Richard (2000) shows that diversity effectively adds to the firm value, and enhances the firm’s competitive advantage. Grossman and Helpman (1994) and Ederer et al. (2007) state that the ability to innovate depends greatly on the quality of human capital. In compliance Bassett-Jones (2005) and Mulholland et al. (2005) emphasize the importance of diversity in human capital as this facilitates those businesses which innovate and strive for continuous

improvements in products and/or processes by enhancing creativity and problem-solving capability and enhance customer relations.

28 The differences in defining performance may be an explanation for the conflicting empirical findings of studies analysing the relationship between diversity and firm performance

(Williams and O’Reilly, 1998). Finance literature for example has developed several metrics to capture the performance of firms (Jensen, 1968; Fama and French, 1993 and Carhart, 1997) but these quantitative metrics vary greatly from performance measures from other research disciplines. The psychology area uses a more qualitative approach for example (Pelled,

Eisenhardt and Xin (1999). While literature on the measurement of firm performance contains great disparity the same applies for literature on measuring innovation at the firm level. The following section deals with this topic.

3.9

Measuring firm’s innovativeness

The level of innovativeness a firm possesses is hard to grasp. Many authors have tried to define the actual meaning of innovativeness and alongside the definition proposed tools for measuring a firm’s level of innovativeness. This paragraph of the theoretical reflection will focus on literature on innovativeness at the firm level.

The most commonly used measure of innovativeness of a firm is the per cent of revenue from new products. The measure is appealing to use for many reasons. It is quantitative, it implies a rate of generation, it is relatively easy to measure and appears easy to understand (Shapiro, 2006). If revenue as measured by the accounting system can be associated with specific products, the measure can be derived without difficulty from the Enterprise Recourse Planning (ERP) system. The measurement allows a consolidated view of the entire

organization including its separate divisions and business units. It is a measure associated with companies considered innovative, such as 3M and DuPont, and is one used by many

companies in their pitches to Wall Street (Kanter et al., 1997). The difficulty regarding this measurement lies in the definition of “new” or “new products”. Shapiro states that the measure is a good measure at its core and can be made to be a superbly useful measure if these two flaws are addressed: the difficulty of applying it consistently, and the bias toward one type of innovation over all other.

29

3.10 Innovation in service industries

As described earlier innovation measurement in academic literature has been conducted in various ways and mostly focussed on product innovation in manufacturing industries. As this study aims at measuring innovativeness in the Dutch marketing and communication segment, being a service industry, a different measurement method is needed. In order to establish a comprehensive measurement set for this industry this section clearly describes the innovation process within service industries.

Within service industries the products are not always perfectly “formatted” and codified. Each service transaction may be considered as unique as far as it is produced on demand (tailor-made) in interaction with the client or as a response to specific, not standardisable problem, and in different environments (Sundbo and Gallouj, 1998). This results into the innovation process in service industries being to a large degree an interaction process, both externally and internally. The external interaction mainly relates to the customers. In service production the customer’s satisfaction with the total encounter (not only the core service delivered, but also the circumstances of delivery) is of crucial importance. Customer satisfaction is of greater importance than the issue of a new core service. Customer satisfaction, in terms of service quality, has thus been more important to service firms than innovation. Also at the internal level the innovation process within service industries is an interaction process. Innovation is generally an unsystematic, collective process in which employees and managers participate in different interaction patterns at the formal and the informal level (Sundbo and Gallouj, 1998). According to Sundbo and Gallouj (1998) the organisation of innovation is differentiated and the various patterns can be observed in different types of firms. In general, service firms have not been very successful at organising the innovation process in a formalised and systematic way and learning from the process, not even when it is regarding the external interaction with customers (Edvardsson, Haglund and Mattson, 1995).

Sundbo and Gallouj(1998) in their study examine whether an innovation systems exist in service industries. They define the concept, innovation system, as a general pattern that can describe the innovation activities in a sector. The existence of a pattern means that certain elements are determining the innovations and the development of new ideas and innovations and their diffusion follows certain paths. When repetition of common characteristics of the pattern occur it can be referred to as a system. Such a system can be institutional or loosely coupled.

30 process follows certain patterns and therefore this process can be tracked. The

interrelationship between the actors often follows a certain pattern with long lasting relations and co-operations and often the relationship is formalized through contracts or well-known norms for co-operation, therefore the system can said to be institutionalized. Generally accepted norms for the relationships and interactions together with fixed positions in the system contribute to the institutionalization of the system.

A loosely coupled innovation systems is a composition of actors, trajectories of development and some forms of behaviour that are common to the sector in case. The constellation between the actors, the trajectories and the behavioural forms is not fixed and thus may take various forms. In contrary to an institutionalized system there are no fixed norms for

behaviour and no relations that everybody in the sector knows and follows. The co-operation with external actors is lower in comparison to an institutional innovation system, however firms with a loosely coupled innovation system relate to the external actors as they are important providers of input or purchasers of the output from the innovation process (or both) or are competitors. The diffusion process does not follow a straight line; it is a complex process with many informal and often in-observable elements. The loosely coupled innovation system cannot be theoretically understood from a coherent, explanatory model as the

institutionalized system due to the loose coupling of all the elements and the non-fixed behavioural patterns. A loosely coupled innovation system often characterizes a situation with fierce competition and weak common push elements such as a common scientific or

technological basis of the production (a trajectory). Such a situation calls for a strategic game approach towards the market and little co-operation with other actors outside the firm. Sundbo an Gallouj (1998) conclude in their study that within service industries innovation systems do exist. There is some coherence within service industries when it comes to innovation trajectories, a limited number of innovation patterns is repeated across different service industries, so therefore one can speak of the existence of a system. Due to the lack of strong coherence and the existence of several innovations patterns the authors state that it is only a loosely coupled system.

Furthermore, Sundbo and Gallouj (1998) state that different types of service firms entail different types of innovation patterns, resulting from a set of internal and external driving forces referred to as the components of these innovation patterns. A full explanation of all these driving forces lies beyond the scope of this study, however it is important to explain the external driving forces that influence the innovation pattern that fits the marketing and communication industry. Sundbo and Gallouj (1998) divide the external forces into

31 the innovation activities of a firm can contribute to the reinforcement of a given trajectory. Trajectories can be described as ideas and logics that are diffused and adopted through the social system, being a nation, an international network, a professional network, etc.. The diffusion occurs through many actors that are often difficult to identify. The most important factors however are not the actors involved but the ideas and the logic behind these ideas. Sundbo and Gallouj (1998) identify five types of trajectories, only the relevant trajectory for the Dutch marketing and communication segment will be discussed here, being the service professional trajectory. It entails the methods, general knowledge and behavioural rules (ethics) that exist within the different service professions.

The innovation pattern that fits the marketing and communication segment the best is the pattern described by Sundbo and Gallouj (1998) as the service professional pattern. This pattern, introduced by Barcet, Bonamy and Mayere (1987), characterizes the professional knowledge service firms. In general the firms are medium sized and devoted to knowledge intensive business services, often referred to as ‘knowledge-intensive business suppliers’ (e.g., Miles et al., 1995; Den Hertog,2000). The firms do not really sell product-services, but competencies, abilities to solve problems in different expertise areas. Formalised structures dedicated to innovation do not exist in this pattern. The innovation trajectory is of the service professional type, as is described above.

Figure 2. The service professional innovation pattern, source: Sundbo and Gallouj (1998)

Within this pattern the innovation process is a collective process in which all professionals are supposed to participate. The pattern is a flexible one ensuring that the firm is able to answer quickly to the market signals and the members are able to cross the individual ideas

32 process and the absence of enterprise projects.The main driving force for innovation is the individual expertise and competencies of the organizational members which correspond to the service-professional trajectory. The interface with the customer is an important locus in the service professional pattern. The client plays an active role in the innovation process and the ad hoc mode of innovation, without being exclusive, is highly present. Ad hoc innovation regards the implementation of service products that require an adaption or translation to client specificities. The implemented ‘product’ is essentially coproduced with the clients as an interactive problem solving process, geared to the contexts of a specific client. The originated innovations are augmented by formalizations of implementation experiences . By making use of ‘appreciative theorising’ the repertoires of implementation towards new users are enhanced (Sundbo and Gallouj, 1998). In their study Sundbo and Gallouj (1998) show through an analysis of the different steps of such ad hoc mode of innovation that the steps of production, selling and innovation take place simultaneously or are merged. The starting point of the innovation process is the client’s ‘problem’ (in its concrete sense). The service production process later on becomes an innovation process and ends with a formalisation step. This formalisation step is achieved without the client’s participation and aims at going through the problem and the innovating final solution again as well as formalizing and modifying them in order to re-appropriate some of their components and to capitalize them in the memory of the firm (eg. paper, software, IT-files, etc.).

Ad hoc innovations are a product of customer interface and depend upon the nature and the components of this. With regard to this sparring type interfaces are most propitious for the creation and success of this form of innovation, since they help to assure a better

understanding and acceptance (legitimacy) of the innovation. Sundbo and Gallouj state that the opportunities for ad hoc innovations increase with the size of the service provider and that of the client. Successful implementation of ad hoc innovation also depends upon the quality of the professionals in the service firm participating in the interface.

De Jong et al. (2003) also define sectoral differences when it comes to innovation in services. Their findings are comparable to Sundbo and Gallouj (1998). The sector that fits the

33 innovations. Usually the co-workers are responsible for innovation in their daily work and therefore no separate R&D-departments exist. Typically, these firms operate in a business-to-business environment, having only a few customers with relatively long client contact times (Silvestrou et al., 1992). Usually these firms make considerable adjustments in their services to meet customer needs and therefore ‘client-led-innovation’ and ‘innovation through services’ are dominant types of innovation in these sectors. In the case of client-led-innovation the service firm is responding clearly to the needs articulated by its clients. This innovation classification is comparable to the type of innovation occurring in the service professional innovation pattern by Sundbo and Gallouj, described above. Innovation through services is a more complicated type of innovation and mostly found in business-to-business industries. In this type, service firms influence the innovation process taking place within the client firm. The service firm may provide knowledge and/or resources that support the innovation process in various ways. Despite the inputs most, if not all, of the innovation process takes place at the client’s site (de Jong et al., 2003). The direct interaction between the provider of a service and the customer and the absence of involvement of a back office means that a more systematic evaluation of the innovation trajectory by an external party cannot be implemented (Dolfsma, 2004). This also ensures that measuring innovation becomes more difficult.

34

4. Research question and conceptual model

The main research question to be answered in this study, as formulated in the introduction, reads:

What is the effect of human capital diversity on the innovativeness of Dutch marketing and communication agencies?

The focus is thus on the relationship between human capital diversity and firm innovativeness. Moreover, two additional factors are examined whether or not they influence this relationship. First the potential influence of the fulltime versus part-time employee ratio is investigated. Secondly, the study explores if the share of high educated employees within the organization may affect the relationship between human capital diversity and firm innovativeness. Human capital diversity in the realm of this study is determined by a combination of four underlying variables. The variables gender, age and nationality are combined into surface-level diversity (Harison, Price and Bell, 1998). Human capital diversity is completed by the fourth variable educational background diversity. Firm innovativeness is constituted by three underlying variables which have been discussed in the innovation typology section namely service innovation, process innovation and a third variable, targeted at measuring innovation containing a high degree of newness, labeled newness of innovation. A visual representation of the study is shown in the conceptual model below.

35 The concept of surface-level diversity in relation to innovation asks for a brief explanation as this is new within the innovation literature. In the continuation of this study the diversity variables age, gender and nationality will be aggregated in terms of their effect on firm innovativeness. As far as it is known this study is the first to investigate the effect of surface-level diversity (as a combined variable of age, gender and nationality diversity) on innovation. On basis of the existing literature on the effect of these diversity variables on innovation no ranking order in terms of the power of the effect of each variable can be made as literature does not discuss this. Therefore, in this study each underlying variable of surface-level diversity has an equal influence on the formation of the surface-level diversity variable and thus the also on the dependent variables.

Apart from the surface-level diversity variable the effect of diversity in educational background of the employees on the firm’s innovativeness will be tested. Together with surface-level diversity these two variables constitute human capital diversity in the realm of this study. The dependent variable firm innovativeness consists of three underlying sub-variables being service innovation, process innovation and a variable targeted at measuring innovation with a high degree of newness labeled newness of innovation.

4.1

Hypotheses formulation

As stated before, innovation is an interactive process and diversity among those who interact promotes the innovation process since diversity affects the way knowledge is generated and applied in the innovation process . Thus human capital diversity should generally have a positive influence on innovation however, high levels of diversity might create conflict and slow down the innovation process (Ostergaard, Kristinsson and Timmermans, 2009). A trade-off exists between diversity and the commonality across individuals (Cohen and Levinthal, 1990).

Based on previous literature the hypotheses in this study are aimed at testing if surface-level diversity and diversity in educational background constitute a positive influence on

innovation. All hypotheses formed are explanatory hypotheses, which means that the existence of or a change in one variable causes or leads to a change in the other variable (Cooper and Schindler, 2006).

36 age, gender and nationality are to be discussed separately in the light of the expected influence on innovation.

Literature suggest that the influence of gender diversity on innovation depends heavily on contextual factors. While in male-dominated professions gender diversity most likely will have negative impact on innovation, gender diversity in a more gender-balanced profession

positively influences innovation (Joshi and Roh, 2009). The sample for this study shows a fairly equal balance between male (53%) and female (46%) employees. Therefore a positive effect of gender diversity on innovation is expected.

The daily work in the marketing and communication business entails creativity as a

centerpiece. Literature on age diversity in relation to innovation suggests that age diversity is more beneficial for more complex tasks (Wegge and Schmidt, 2009; Wegge et al., 2008) and those requiring creative behavior (Choi, 2007; Thornburg, 1991; Triandis et al., 1965). Therefore again a positive effect of age diversity on innovation in this study is expected. Although the research on nationality diversity has been scarce so far, some studies reported benefits for nationality diversity for innovative output. McLeod et al. (1996) and Watson et al. (1993) showed that the quality of solutions presented or ideas generated improved alongside higher ethnical diversity. In a study among Irish manufacturing firms by McQuirk and Jordan (2012) a positive relationship was demonstrated between nationality diversity among the labour force and product innovation.

There is a priori no reason to believe that any type of diversity constituting surface-level diversity has a greater impact on innovation than the other as no literature reports on this. However, this study is among the first to investigate the effect of surface-level diversity on innovation within service industries and therefore further research should generate more insight into this. Studies on the effects of the three types of diversity all report a positive influence on innovation and in line with the findings presented above a positive effect of surface-level diversity on innovation is expected represented in the following hypothesis:

Hypothesis 1: Surface-level diversity positively influences innovation

37 In line with these studies the following competing hypotheses are formulated with regard to the effect of HCD on more radical innovation:

Hypothesis 1a: Surface-level diversity negatively influences innovation entailing a high degree of newness Hypothesis 1b: Surface-level diversity positively influences innovation entailing a high degree of newness

Ostergaard, Timmermans and Kristinsson (2009) in their study showed significant

relationships between the level of educational degree and innovation performance as also for the diversity in educational disciplines and the likelihood to innovate. In line with these finding the following hypothesis is established:

Hypothesis 2: Educational discipline diversity positively influences innovation as well as innovation entailing a high degree of newness

38

5. Methodology

In this methodology section the design of this research will first be outlined. Subsequently the questionnaire that was distributed among the Dutch marketing and communication agencies will be discussed. Moreover the sample, methods of data collection and the research

approached are described. The data and measurement section further elaborates on the underlying statistical analysis that are conducted in order to test hypothesized relationships.

5.1

Research design

This study uses a quantitative approach in order to test the hypotheses formulated above. The data is gathered through a questionnaire made available online. As the complex nature of the marketing and communication research causes some ambiguity around specific attributes of the study this quantitative approach is supplemented with additional qualitative data. Several experts from the field were contacted to give their view on the actual definition of these key instruments as “service” and “innovation” in order to formulate a universal meaning that fits the industry as a whole. On basis of this of this formulation the questionnaire was formatted ensuring that all the participants had the same meaning towards the variables in the study. As well the quantitative aspect in the form of the questionnaire as the qualitative outcome from the industry experts are to be further discussed below.

5.2

The questionnaire

A questionnaire is drafted in order to test the hypotheses formulated above. This

questionnaire is based on the Community Innovation Surveys (CIS) in 2006. The CIS are a series of surveys conducted by national statistical offices throughout the European Union and aimed at retrieving information about a large set of organizations active in different sectors. The format of the survey is in line with the guidelines provided by the OSLO Manual. In order to reduce the chance of a low degree of correspondence between the participants perception and the actual meaning of the question each type of innovation is explained into detail, clearly delineated and supplemented with practical examples alongside each