Technological Forecasting & Social Change

Transcription

1 Technological Forecasting & Social Change 79 (2012) Contents lists available at SciVerse ScienceDirect Technological Forecasting & Social Change Revisiting incubation performance How incubator typology affects results José L. Barbero a,, José C. Casillas b,1, Alicia Ramos b,2, Susana Guitar c,3 a Universidad Pablo de Olavide, Ctra. Utrera Km. 1, Sevilla, Spain b University of Seville, Departamento de Administración de Empresas and Marketing, Avda Ramón and Cajal, 1, Sevilla, Spain c Director of Innovation, Technology and Business, Andalusian Innovation Regional Ministry, C/ Albert Einstein 4, Sevilla, Spain article info abstract Article history: Received 12 March 2010 Received in revised form 14 November 2011 Accepted 6 December 2011 Available online 4 January 2012 Keywords: Incubation Types of incubators Performance Growth Relevant literature has discussed incubator versus nonincubator firm performance and the characteristics of different types of incubators. However, it has not considered assessing performance individually by archetype. We studied company performance in four archetypal incubators. Performance measures used fall into five categories drawn from literature: a) firm growth b) participation in R&D programs c) Input R&D d) Output R&D e) Employment generation cost. We find there are significant differences in three of the five performance categories among incubator types. We assess the performance of each one by determining whether the objectives for which each was created are met. Private, basic research and University archetypes meet their objectives. Regional development archetype does not Elsevier Inc. All rights reserved. 1. Introduction A vast majority of incubator literature deals with the differences in performance between incubator firms and firms located outside science park protection [1]. Literature is inconclusive as we can find results supporting both claims: the value added of the incubators and those that claim they provide little value added [1-3]. An incubator is justified based on superior innovation performance; as studies have been inconclusive, we argue that performance differs according to the type of incubator. We aim to study the differences in performance of each type of incubator in order to avoid comparing apples to pears [4]. In our research, we review the relevant literature on types of incubator [4-7] and define the four most common types used in Spain which are similar to those found in other Latin countries [6] and those archetypal incubators described in current literature [8]. The four archetypes are: basic research, university, economic development and private incubator. Based on literature and based on local legislation we determine what are the objectives each archetype is created for. Subsequently, we review performance measures used in literature to assess incubator firms and conclude there are five categories mostly used for research purposes: firm growth [2], participation in R&D programs [9], input R&D [1,10], output R&D [1,10] and employment generation cost [43,44]. We do not use survival as an indicator as this has been widely criticized [10]. The purpose of our research is to establish whether there are significant differences in the performance of the four archetypes of incubatee firms and whether each type individually meets the objectives for which it was created. We discover striking differences in three of the five performance set of measures: participation in R&D programs, input R&D and output R&D. Specifically, we find that university incubator firms outperform the other three types in the participation in R&D programs, private incubator firms Corresponding author. Tel.: addresses: jlbarnav@upo.es (J.L. Barbero), casillas@us.es (J.C. Casillas), aliramos70@hotmail.com (A. Ramos), susana.guitar@juntadeandalucia.es (S. Guitar). 1 Tel.: ; fax: Tel.: ; fax: Tel.: /$ see front matter 2011 Elsevier Inc. All rights reserved. doi: /j.techfore

2 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) outperform basic research and economic development incubator firms in one of the input R&D indicators (qualified scientists and engineers, hereinafter, QSE) and basic research incubator firms outperform economic development and university incubators in one of the output R&D indicators (introduction of new products and services). In terms of goals, we find that economic development incubators do not meet the objectives for which they are created. The other three archetypes meet the objectives, however, basic research and private archetypes meet them successfully, whereas university incubators only meet them satisfactorily. Literature has traditionally analyzed incubation on the basis of case studies. Some authors have argued that incubators are analyzed in an atheoretical and unempirical fashion [1,11] and that analysis of perceptual data of a small number of parks [10] is inappropriate. Incubators are mainly publicly funded and obtaining data is difficult per se. We advocate the implementation of empirical studies as an important source of business incubation research. Hence, we test our results using one of the largest samples in incubation literature, as we work with a sample which includes seventy incubators. In summary, the literature gap we address is the study of how different types of incubator perform based on whether the archetypes meet the objectives for which they were set up. Focusing on the literature gap, in the next section we start by reviewing literature on the heterogeneity of incubation, the types of incubator and performance measures. In Section 3, we review the research context by describing incubation in Spain and describing archetype objectives. Section 4 discusses the results obtained and finally, Sections 5 and 6 close with a clear set of conclusions, policy implications and a set of interesting areas authors should focus on in the future. 2. Theoretical underpinnings 2.1. Incubator heterogeneity Many authors argue that we need to take into account the specific features of incubators in order to assess performance. This in turn would lead to an improved incubator policy response by governments. From the beginning literature has discussed heterogeneity of incubation. Schwartz and Hornych [12] suggest that studies of incubatees have been conducted in diversified sectors offering poor incubator performance results, however, studies on sector specialized incubators can demonstrate better performance results. The authors used a case study of an incubator in Halle (Germany) to show evidence of the benefits of sector specialization. There are two benefits derived from sector specialization: 1) high quality advisory services, premises and equipment, which benefit both the incubator (cost reduction) and the incubatees (quality of advice, tailored premises) 2) image effects of the location (media presence, positive word of mouth). The results do not support advantages in terms of inter-firm and university firm networking activities. Aerts et al. [13] confirm the benefits of specialization. The downside of specialization is argued by the author, as a concentration strategy which increases vulnerability if the sector of specialization recedes. Managed science parks versus non-managed parks is another category found in literature [14]. A managed park has a full-time manager whose main function is to oversee the park. Non-managed parks do not employ a full-time manager at the park. The authors [14] concluded that managed parks have a higher closure rate and host firms are less established and resource poor. They also use more facilities and central services than non-managed parks but appreciate the role of the park manager. Incubator type is another area where specialization can offer up evidence of specific characteristics of incubators that should be taken into account in order to deepen our understanding of incubator reality. Allen and McCluskey [5] are one of the first group of researchers to indicate the heterogeneity of incubators and the importance of incubator types. The authors suggest a relationship between incubator type and public policy. In order to lead to improved public policy decisions, traits of a specific incubator category should be considered. Aernoudt [4] points out that there are a large number of incubators in the world but they possess uneven quality. He proposes that in order to evaluate the quality we need to fine-tune the concept in order to avoid comparing apples to pears. Aernoudt [4] concludes that the evaluation of incubators needs to take certain differences into consideration as incubator archetypes have different missions. The definition of different types of incubators can help us uncover a heterogeneous reality Types of incubator Classifications of incubators differ depending on geography. Literature studying types of incubators in the United States [5] shows differences between the types of incubator in Europe [6,15]. Differences arise as incubation has been recognized as distinct in both continents [4,16]. Even in Europe there are differences in the types of incubator used as business policy instruments. Grimaldi and Grandi [6] and Von Zedtwitz and Grimaldi [15] offer a classification based on incubators in Italy. In spite of different types, depending on the host country, a consensus on the types of business incubator has been demonstrated in literature [4]. This is essential in order to make progress in the assessment of differences in the incubator type performance. A table summarizing the different classifications by authors is shown (Table 1). Pioneer research into incubator classification was conducted by Allen and McCluskey [5]. The authors developed a continuum distinguishing types of incubators depending whether value is added through real estate or through an enterprise development program. Six types of incubator emerge: for-profit property development, not-for-profit development corporation, academic, forprofit seed capital, hybrid and corporate. Aernoudt [4] distinguishes five types of business incubators in accordance with the main philosophy, objectives and sectors involved. Mixed types were first born as tools for reconversion, offering services both to low and high tech incubatees. Regional

3 890 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) Table 1 Description of incubator classification by author. Authors Types of incubator in the classification Variables used to generate classification Allen and McCluskey [5] Sources of value added For-profit property development Not-for-profit development corporation Academic For-profit seed capital Hybrid Corporate Aernoudt [4] Main philosophy Objectives Sectors involved Mixed Regional development Technology Social Basic research Carayannis and Von Zedtwitz [8] Von Zedtwidtz and Grimaldi [15] Grimaldi and Grandi [6] Clarysse et al. [16] Becker and Gassman [7] Competitive focus Strategic objectives Competitive scope Strategic objectives Private/public nature Institutional mission Industrial sector Location Market Origin of ideas Phase of intervention Sources of revenue Services offered Management team Level and complexity of activities performed Heterogeneity of resources deployed Source of technology Type of technology Regional Business University Independent commercial Company internal Virtual Regional business University Virtual Independent commercial Company internal Business innovation centre University business Corporate business Independent business Low selective model Supportive model Incubator model Fast profit Leveraging In-sourcing Market development types focus on promoting regional economies and improving regional competitiveness. Technology incubators aim to develop technology oriented firms. Social incubators' main objective is to support people with low employment capacities. Basic research incubators concentrate on incubatees involved in basic research. Carayannis and Von Zedtwitz [8] agree that a general classification of incubator types would yield similar results to Aernoudt [4] as, they argue that incubation archetypes are regional business, university, independent commercial, company internal and virtual incubator. Von Zedtwidtz and Grimaldi [15] found evidence in Italy of a classification of incubators based on Porter's four elements of competitive scope (vertical scope, segment scope, geographical focus and industry focus) and the strategic objective (for-profit, not-for-profit). The five incubator types that emerged were regional business, university, virtual, independent commercial, and company internal. Regional business and university are of a not-for-profit nature, whereas independent commercial and company internal are private. The virtual type is characterized by a for-profit nature, a lack of physical location and preference for online services. The independent commercial type invests in high-tech sectors, is located in industrialized areas and is managed by individuals committed to their ventures who rely on private funding. Company internal types host ventures whose core is related to the parent company and receive funding. Grimaldi and Grandi [6] introduce yet another classification of business incubators which comprises five main types. Most of the types recur throughout literature as their classification includes business innovation centers, university, corporate business and independent business incubators. These centers are incubators set up by a European Commission initiative to provide basic incubation services. The authors placed each of the five incubator types in a spectrum with two models of incubation on each side, model 1 and model 2. On one side resides model 1, which is mainly characterized by its public nature. Model 1's main competencies relate to the capacity for the provision of physical assets at low prices and to the facilitation of access to funding and expertise. Model 2 shows incubators of a private nature with a short-term time orientation providing incubatees with finance and more intangible and high value assets such as networks. University incubators are situated in the middle of the continuum, as they receive public funding but have access to technological knowledge and academic networking. Clarysse et al. [16] found three types of incubator models based on the level and complexity of activities performed and heterogeneity of resources deployed. The low selective model objective is to create as many start-ups as possible with no technological focus

4 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) Table 2 Summary of incubator performance measures used by literature. Performance model category Author Measures used Complex models with large number of measures Bigliardi et al. [23] Patrimonial structure Internal development Repercussion in the territory Economic and financial aspects Human resources and technical-scientific productivity International and interregional relationships Mian [2] Goal approach (4 measures) System resource approach (7 measures) Stakeholder approach (4 measures) Internal process approach (5 measures) Complex models based on assessment of Input & Output R&D measures Westhead [1] Input R&D: - Proportion of qualified scientists and engineers (QSE) - Radical new research - R&D spending/sales - Gross R&D investment/sales Output R&D: - Number of patents - Introduction of new products or services. Colombo and Delmastro [9] Input R&D: - Researchers/total employees - Employees with graduate degree/total employees - Firms involved in EU R&D projects (%) - Firms that have purchased R&D services (%) - Number of PC per firm - Number of workstations per firm - Firms with LAN (%) Output R&D: - Firms with patent activity (%) - Firms with copyright (%) Simple models with several measures Siegel, Westhead and Wright [10] Survival rate Employment growth R&D activity Allen and McCluskey [5] Occupancy Jobs created Firms graduated One measure models Aerts, Matthyssens and Vandenbempt Survival rate [13]; Aernoudt [4] Peña [22] Growth and lacking private capital. The supportive model limits the number of ventures, employs a larger multi-disciplinary team and provides services such as IP, patent and license negotiation with industry. Finally the incubator model can count on experienced professional staff, has access to private capital and has a narrow focus on a certain type of technology. Becker and Gassman [7] introduce a classification of corporate incubator types, which is found on the private side of the incubation continuum. The authors identify four types of corporate incubators in technology corporations based on an incubator's source of technology and the type of technology (core/noncore). The four types are fast-profit, leveraging, in-sourcing and market incubators. Insourcing incubators are characterized by adopting parent core technology which is externally acquired and by achieving higher survival rates. Bergek and Norrman [17] describe an incubator model based on three variables: selection (according to the strategy selected), business support (scaled from strong intervention to laissez-faire) and mediation of regional or national innovation systems; technological or sectoral innovation systems). We have summarized the above literature which deals with classifications of incubator types but there is also some research by authors who have implicitly studied specific incubator types [9,18]. In this sense, new types of incubators are still emerging. Bollingtoft and Ulhoi [11] argue for the networked incubator as a new incubator model. This model is characterized by a bottom up approach in which new ventures develop and manage the incubator. Ventures in the incubator engage in business networking due to complementary skills. Nowak and Grantham [19] propose the virtual incubator which is characterized by the distribution of resources through the internet, a focus on strategic alliance formation, a key role played by the private sector and work in conjunction with physical incubators. Carayannis and Von Zedtwitz [8] propose the G-R VIN (Glocal Real & Virtual Incubator Network) incubator type which takes best practices from the five incubator archetypes Incubation performance measures Incubator performance measures is a widely discussed issue and has become one of the most controversial points among authors; literature broadly agrees that no clear consensus exists on what constitutes an appropriate measure of performance [20,21]. The

5 892 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) problem is often compounded as objective and direct performance indicators are very hard to obtain [13]. In this section we review the different performance measures used in literature. A table summarizing such literature has been included (Table 2). Some authors have produced complex models for the evaluation of an incubator [2,23]. The models are judicious and consider the issues at stake. Bigliardi et al. [23] propose an evaluation model based on six elements: patrimonial structure, internal development, repercussion in the territory, economic and financial aspects, human resources and technical-scientific productivity and international and interregional relationships. Each element is analyzed through an average of six indicators. The evaluation of six Italian incubators follows the model discussion. Mian [2] proposes a complex performance framework based on four approaches which are goal approach, system resource approach, stakeholder approach and internal process approach. We think that complex models suit the evaluation of a limited number of cases, however, they are not apt for the evaluation of studies comprising larger samples, such as those considered by Allen and McCluskey [5] and Aerts et al. [13] which comprise one hundred and twenty seven and one hundred and four incubators respectively. Additionally, for public policy purposes, the compilation, analysis and decision-making process can become gruesome. Other studies use an array of input and output R&D indicators [1,24]. Westhead's perspective on performance [1] is one of the most followed. For input R&D, he uses the proportion of QSEs employed in a firm [25], primary research thrust as radical new research, and some financial indicators measuring R&D intensity such as R&D spending or gross R&D investment as a percentage of total sales revenue. For output R&D, he uses the number of patents [26,27] and the introduction of new products or services [24,26] either for existing clients or for new markets. Colombo and DelMastro [9] agree with Westhead [1] as they use a similar set of indicators, namely the number of researchers, patents and copyrights, as well as other different indicators such as the adoption of advanced information and communication systems; participation in R&D programs promoted by the EC; establishment of formal collaborative agreements; and lastly, firms that have purchased R&D services from universities among others. Siegel et al. [10] suggest the use of three key indicators, namely, survival rate, employment growth and R&D activity. Many incubation studies use a venture survival rate either as a sole indicator [13,27] or as part of a set of indicators [2,22,29]. Phan et al. [21] believe the measure lacks construct validity and creates an endogeneity problem as incubators are designed to maintain incubatees alive. Other authors have used growth as a measure of performance. Growth is represented by employment growth and sales growth [2,22]. Other authors [22] even use profit growth as well. We do not believe profit growth is a valid indicator as growth literature proves measures should be employment and sales growth [30,31] given that high performance firms, in the initial stages, often lack profits. The cost per job measure is less used in literature [43,44] but more used in economic development programs [45,46]. The reason being is incubators are tools for economic development, cost of job creation should be compared with other instruments of economic development. The measure is often used in the U.S. [46]. Other measures of performance less frequently used are backward patent citations for university incubates [32], the adoption of advanced information and communication technologies as well as participation in R&D programs promoted by the EC [9]. Some authors have correctly pointed out that the evaluation of science parks depends on the legal form [23,33]. Some parks are for-profit entities while others are not-for-profit. For-profit can strictly be evaluated through financial indicators [23]. However, in Europe the nature of incubation is mainly not-for-profit [4]. Mission statements tend to differ between incubators too, and some authors attribute such heterogeneity of mission statements to the lack of agreement on performance evaluation [23,34]. Bergek and Norman [17] toy with the idea that incubators should be evaluated based on the goal of the institution. Although we agree with the authors, we think there can be many levels of goals. Every incubator is unique and different. Incubators of the same type but operating in different sectors may have similar goals. We believe the correct grouping for goals is the type of incubator: incubators belonging to the same type typically have similar goals. Finally, few researchers have analyzed incubator type performance [4,5]. Aernoudt [4] compares the type of incubator considering performance measures. Metrics of performance used by the author are survival rate and job generation. In this sense, the author is one of the few to provide performance figures for some types of incubator. Survival rates for mixed, economic development and technology incubators are 87%, 86% and 90% respectively. In summary, assessing venture performance should ideally be attained through a complex model accounting for all the relevant influences. However, complex models are valid for case studies, as research on complex models typically uses a sample lower than seven incubators. For empirical research produced with large samples, it is advisable to use a set of indicators which would facilitate analysis and conclusion drawing. We propose a group of indicators widely used in literature. They are based on input R&D and output R&D as these are the measures most used in literature either within a complex model or by themselves. Our output R&D measures are employment growth, sales growth, patents generated and introduction of new products and services. For input R&D we use QSE and R&D intensity, R&D spending and gross R&D investment as a percentage of total sales revenue. Venture survival will not be used, as we agree with Phan et al. [21] that it presents an endogeneity problem. 3. Methodology and sample 3.1. Incubator types in Spain To create a classification of incubator types in Spain, we will start by describing what the terms cluster, science park and incubator mean, as literature has highlighted problems with these definitions [21,35]. A cluster is defined by Porter [36] as a geographic concentration of interconnected companies, specialized suppliers, service providers, firms in related industries and

6 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) associated institutions in specific fields that compete but also cooperate. A science park includes three components [37]: a real estate development, a program of activities for technology transfer and a partnership between academic institutions, government and the private sector. Science parks tend to house incubators [37]. An incubator is an entity which provides small businesses with resources [5] that improve their chances of foundation and survival [38,39]. The objective of the incubator is to improve the probability of survival of the firms supported [40] and accelerate their development [18,41]. Our definition of incubation is sufficiently ample as to host both public and private initiatives whose purpose is to assist the development of small companies. As discussed, country disparities can be observed, as types applicable in one country are not applicable in another country. For our purposes, we cannot blindly follow a classification which has been established by research performed in another country because policies introduced in different countries will lead to different incubator type outcomes. In any case, we can reclassify the Spanish case into different archetypes introduced by incubation literature. To create a classification of incubator types in Spain, we need to exclude some clear types operating in other countries based on features of the Spanish innovation system. This will help us to come up with a classification of the incubator types most common in our territory. In Spain, there is no clear cut type of technology incubator as this type of firm tends to be hosted in incubators located at a university, in a technology cluster created by a policy decision and even at business incubation centers, however, features among this type of incubator differ widely. Another Spanish feature is the lack of social incubators, although this does not imply that the country does not promote business incubation for people with lack of employment opportunities. It is just that private and public agents have not created an incubator type with this main objective in mind. There are certainly some virtual incubators in Spain, however, they are not common. This type of incubator has not proved popular in Spain, probably due to a lack of promotion through governmental initiatives, as officials have promoted other types. In order to develop an incubator classification, we looked into the archetypes' objectives. Objectives are very important to incubator literature [6,8,42] as such is the most widely used and sometimes the most important criteria [4] to distinguish between archetypal business incubators: Aernoudt [4] argues the criterion is best as it is closer to economic reality [4]. Some incubators will follow public policy objectives such as regional employment and GDP growth or public objectives such as employee retention, innovation capacity building or stock market valuation [8]. The determination of the incubator objective will allow us to determine later how successful each archetype is when we determine incubator performance. Many other variables exist, however, we believe they are a consequence of the objectives and by defining the objectives, the rest of variables and attributes will be set. Our research team had access to information on incubators in Andalucia, a Southern region of Spain comprising 20% of the country's population. The region has received large amounts of funds from the European Commission throughout the years, as it was considered a region Objective 1 under the European Commission Decree. In addition, the region has made a tremendous effort to encourage entrepreneurship which has fostered a strong infrastructure oriented to the promotion of entrepreneurship and this has resulted in the network of existing incubators in the region. The innovation system in Andalucía is defined by decree 254/2009 of May 26th. Decree 254/2009 includes a detailed definition of the innovation agents in the system either of a public or a private nature. The decree includes several types of incubators as part of the innovation system. A discussion follows regarding these types A typology of Spanish incubators and their objectives Four types of incubators are listed in the decree. The first type corresponds to the basic research incubator described by Aernoudt [4]. The author describes this type as linking incubation with fundamental research. In this type, technologies developed take the form of intellectual property that can be licensed by commercial partners or exercised by spin off companies [4]. According to the decree, the basic research incubator can be a for-profit or not-for-profit agent whose main strategy/mission is to boost naturally formed clusters. These clusters were reinforced with a specialized incubator to help create future sector leaders. The ventures focused on the production of IP protectable R&D and the generation of patents. Due to such specialization, incubatees in basic research incubators have access to very specific resources such as personnel with a clear knowledge of the sector, specific assets to succeed in such a sector and specialized networks. Schwartz and Hornych [12] have proven the benefits of sector specialization as a strategy for incubation. Also, we believe the strategy develops for basic research incubatees' very specific capabilities. The second type in the decree corresponds to the university business incubator. This has a mixed (public/private) nature as they are dependent on university funding as well as on companies' funds for the transfer of venture generated IP. Incubatees use the non-financial resources of the university's infrastructure extensively in the form of tangible (research equipment and premises), and intangible assets (faculty time, scientific knowledge and contacts). This type has been described at length in literature [2,6,15,18,28,48]. We agree with those authors who believe university incubators do not need to have a focus on technology [6,15] rather than those who strongly believe in a technology focus of university incubators [9,18]. Literature correctly highlights the lack of provision of management capabilities, access to capital and reduction of time to market in this type [6]. The third type of incubator acknowledged by the decree is the economic development incubator [4,15]. Legislation recognizes these are publicly funded entities whose main purpose is the dynamization of economic development through the incubation of SMEs. The decree acknowledges that economic development incubators will provide human and physical assets such as offices, space, equipment, business support, advice and training. Literature describes economic development incubators as focusing on gaps produced by economic differences between areas [4]. The strategic objective of the entity emphasizes less the provision of intangible assets in favor of the provision of tangible ones. The main purpose is the promotion of entrepreneurship in areas

7 894 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) with below average economic indicators. They concentrate on the provision of physical assets at low or zero cost such as daily support, office space, telecommunications access etc. [15]. The private incubator is the final type. The decree recognizes the existence of agents of a private nature operating in the incubation space. It defines companies and R&D departments of companies as part of the Andalucían Innovation System. The incubation efforts of these organizations have a private and corporate nature. They add value through business development [5] and through private financing [6]. Private incubators have a for-profit nature and the objective is to adopt a professionally organized knowledge flow to generate profits from the management of ventures [7]. Private incubators tend to provide capital and access to high value intangible assets such as managerial talent, QSEs and external partners [6] which can lead to fast track to market and to the desired corporate return on investment. Aernoudt [4] argues that incubation is different in each country in Europe but describes how similar regions have similar incubation traits. The author indicates three regions: Anglo-Saxon, German and Latin. In this sense our typology mimics those of other Latin regions, such as the Italian typology described by Grimaldi and Grandi [6] and Von Zedtwitz and Grimaldi [15].Insummary,wecan find a typology of incubators in Spanish legislation that replicates most of the types of incubators described in literature [4,5,7-9,18], particularly the Italian types [6,15]. A summary of the four archetypes found in Spain is found in Table 3. Finally, all the types recognized in the decree comply with the concept of incubator described in Section 3.1 as defined per leading. Such a concept concurs with that of the UK and US leading incubation Associations (UK Business Incubation, National Business Incubator Association) Impact of clientele effect on performance We have described above how incubator archetypes are positioned with a range of services specific to the archetypal. Such positioning generates a clientele effect under which ventures willing to receive services of a specific type become host to that specific archetypal [15]. We have observed that this effect produces an impact on performance through the type of tenant attracted to the incubator. Such an effect was confirmed in interviews with different managers at incubator archetypes. University Pablo de Olavide manager accepted university spin offs as tenants. This type of venture is mostly linked to the research performed by a university department or student final project. The research tends to be intense in innovation. University incubators can help overcome underresourcing by allowing access to university resources and helping obtain access to public funding for innovation. Management lacks experience in launching other ventures or managing a company, however, the incubator does not help by providing this. Incubator manager agrees there is no need to publicize the incubator as tenants are easily captured within the university. Tenants feel the incubator offers what they need. A perfect match between landlord and tenant is generated: incubator set of services is linked to its own objectives. University incubator would offer resources or access to resources to be invested in innovation and expertise to secure IP. Potential university tenants in need of such services end up hosting at the University incubator. These tenants will seek funding for innovation of the IP and new products in the process of development and will boost incubator objectives. The CADE incubator in Jerez is a regional development incubator in a population of over one hundred thousand people. The incubator offers space, telecommunication services and advice. There is no selection process to determine who is getting a place in the incubator. Incubatees are offered a place based on a first come, first served basis. Ventures operate in a wide range of sectors and potential varies. Most ventures operate in mature and difficult sectors with non-innovative products, few operate in novel sectors. Managers do at least have working experience although few have prior experience in launching a venture. Some entrepreneurs have a genuine interest in launching their own business, others are unemployed and end up starting a company as an alternative to unemployment. Table 3 Summary of archetypal incubators found in Spain. Types of incubator Description Authors describing archetype Basic research incubator University business incubator Economic development incubator Private incubator Links incubation with fundamental research. Promotes the generation of technology ventures in strategic sectors. Access to specific sector resources. Technologies developed take the form of IP to be licensed by commercial partners or exercised by spin off companies. Mixed (public/private) nature. Incubatees use extensively non-financial resources part of the university's infrastructure in the form of tangible and intangible assets. Does not necessarily have a technology focus. Publicly funded entities. Purpose is the dynamization of the economic development of the area through incubation of SMEs. Mainly provides physical rather than intangible assets. Private and corporate nature. Purpose is to adopt a professionally organized knowledge flow to obtain profits from venture management. Contribution of capital and high value intangible assets (managerial talent and QSEs, external partners, etc.) Aernoudt [4] Mian [2,18]; Grimaldi and Grandi [6]; Von Zedtwidt and Grimaldi [15] Aernoudt [4] Von Zedtwidt and Grimaldi [15] Allen and McCluskey [5]; Grimaldi and Grandi [6]; Becker and Gassman [7]

8 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) One of the private incubators in our sample is part of a very large listed company with sales of billions of dollars. The company operates in the energy sector which has grown over the last ten years and is expected to keep growing in the medium run. The company has a long tradition in corporate development as one of the top executives was a Partner in a leading consulting firm and implemented a successful and popular corporate development management technique known as the three horizons. The management technique requires companies to treat corporate development as a real option to mature in the short, medium and long run. Each new potential spin off is subject to minimum strategic and financial requirements. From a strategic point of view, each new venture should be strategically aligned to the firm's core businesses. From a financial point of view, the start-up is subject to specific IRR, growth and profit objectives which are closely managed. Ventures are managed by young achievers from internally developed talent. The parent company only accepts projects subject to a potentially high return, intense in IP able to create a competitive advantage. The end result is value added for the parent company, impacted by these tenant and project typologies generated by the firm. In summary, we have observed a strong clientele effect between archetypal incubators and ventures. This effect starts when the archetype defines services to be offered based on the founding by-laws. Ventures in need of that type of service will get to know, through different processes (information through the public administration network, word of mouth by other entrepreneurs, etc.) how the archetype offers the needed services. Entrepreneurs will request to be hosted at the archetype, engaging the matching process. The process ends when the venture leaves the incubator Sample The initial sample was composed of incubators located in Andalucia. Andalucia is the largest region in Spain comprising 20% of the population. Incubators in the region totalled one hundred and eleven. Some incubators had been formed less than four years ago and as such, incubation activity was rather low. We excluded these young incubators and only included those older than four years [20], therefore, the sample was reduced to ninety incubators. Our sample is characterized by weighting more on the public than on the private side, as some authors argue. Aerts et al. [13] describe how 71% of their sample could count on the support of national or regional government funds. Only 29% worked for profit. Aernoudt [4] suggests incubation in Europe is linked to a not-for-profit culture. The research was conducted at the incubator manager level [5,7], as a player with a comprehensive understanding of the entities. The questionnaire was prepared in August/September 2008 in order to be sent out during autumn The questionnaire was previously tested with three incubator managers to understand potential problems arising from questions and any potential difficulties in obtaining the information. A glossary was included in the questionnaire explaining the main entries, especially those of a financial nature. In any case, the main financial concepts matched the definitions used in the Spanish Statistics Institute (INE) technology survey which is widely known. During the pretest we detected that managers found difficulties in financial entries as doubts arose from how wide or narrow the financial entries were. Also managers found difficulties in distinguishing whether the information requested referred to the incubator or the incubatee. The questionnaire concepts were explained in a separate document in order to avoid confusion. Questionnaires were sent to managers via a web application used by the Andalusian Regional Government to exchange data with the Andalucian innovation system agents. To increase the managers' perception of the official nature of the survey, a letter was sent by the Regional Government's Director of Innovation requesting the information. A twelve hour helpline was provided by the Regional Government to answer queries. A one month deadline was set for managers to respond to the questionnaires. Initial response was very low as only 50% of the questionnaires had been completed by the first deadline. A further deadline was granted and took the sample to 90 questionnaires. Twenty questionnaires were not deemed valid as key information needed to perform Table 4 Descriptive statistics for the four types of incubator (mean and standard deviation). Performance measure All types Basic research University Economic development Private A. Growth A.1. Sales growth 8.83% (15.78) 9.88% (18.58) 1.04% (1.28) 2.99% (2.92) 17.19% (12.24) A.2. Employment growth 4.31% (11.94) 5.61% (15.58) 1.24% (1.64) 2.37% (2.23) 4.23% (2.80) B. Participation in R&D programs as a percentage of sales B.1. Participation in domestic R&D programs as a percentage of sales 10.06% (2.42) 2.93% (8.74) 31.02% (40.22) 5.06% (14.48) 2.69% (5.04) B.2. Participation in European R&D programs as a percentage of sales 0.96% (3.22) 0.39% (1.42) 3.03% (5.85) 0.00% (0.00) 0.30% (0.90) C. Input R&D C.1. QSEs employed over total employees 5.38% (14.60) 2.47% (7.15) 8.68% (13.51) 0.68% (2.26) 15.6% (3.25) C.2. Gross R&D investment as a percentage of sales 36.83% (41.84) 35.03% (39.67) 69.54% (41.43) 1.75% (4.81) 24.56% (34.73) D. Output R&D D.1. Number of patents 3.21 (9.01) 4.00 (11.84) 3.82 (7.2) 0.73 (2.4) 2.22 (4.41) D.2. Introduction of new products and services 32.9% (47.31) 51.5% (50.75) 17.6% (39.30) 9.1% (30.15) 22.2% (44.10) D. Cost per job E.1. Cost per job in 4928 (4995) 5537 (5747) 5371 (4366) 4087 (6282) 2986 (1866)

9 896 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) our research was missing. Our final sample reached seventy questionnaires, being one of the largest surveys in incubation literature. By typology, 33, 17, 11 and 9 incubators corresponded to basic research, university, economic development and private incubator type. Descriptive statistics for the total sample and for each type are shown in Table 4. As discussed, public incubators make up a large part of the sample. Public organizations are subject to strict controls and are financially audited by official groups, which are part of the Regional Government. Private incubators, as they belong to larger groups, are audited by auditing firms as part of a listed group or due to shareholders' interest. We do not believe misreporting to have occurred, especially at incubators of a public nature (85% approx.) as supplying misstated information to the Regional Government can be easily tracked later when the incubators are audited. At the date of writing, all the incubators have been audited for the years the information was supplied ( ) Measures Dependent variables Our research focuses on performance differences in types of incubator firms. Literature has traditionally [4,5] argued that incubators differ among themselves and that to extract conclusions on incubation, the type of incubator should be taken into account. We divide incubators into four types which have been widely used in literature as archetypes [4-6,8,15]. The four types which best fit the Spanish innovation system are discussed in the Spanish domestic legislation and have been described at length in the previous section a typology of Spanish incubators ; these four types are basic research incubators, regional development incubators, university incubators and private incubators Independent variables Our independent variable is incubator performance. We assess performance based on measures used by most authors in incubator literature. Performance is measured for different types of incubators in terms of the five following categories: sales growth [2,22], employment growth [2,10,22], participation in domestic R&D programs and participation in European R&D programs as a percentage of sales [9], input R&D [1,9,10,25], output R&D[1,9,10,26] and employment generation cost [43-46]. Growth is measured as the average last four years ( ) of sales and employment growth by firms incubated. Input R&D is measured as the number of QSEs employed in firms as a percentage of total employees and gross R&D investment as a percentage of sales. Output R&D is measured as the number of patents and the introduction of new products and services. Employment generation cost is measured as cost per job which is a widely used measure [43-46]. Cost per job is calculated as the gross cost per job including set up costs and amortization of capital [45]. We do not employ one of the most commonly used performance indicators survival rate as we agree with the critic Phan et al. [21] as discussed in our literature review. In summary, we use five categories of incubator performance (growth, participation in R&D programs, input R&D, output R&D and employment creation cost) and a total of nine performance indicators: sales growth; employment growth; participation in domestic R&D programs and participation in European R&D programs as a percentage of sales; QSEs employed in firms as a percentage of total employees; gross R&D investment as a percentage of sales; the number of patents; the introduction of new products and services; and finally, cost per job. 4. Results In order to identify the differences in the performance of the four incubator type firms we tested them by one-way ANOVA. Results are shown in Table 5. Table 5 Results of ANOVA test for the performance measures. Performance measure Significance levels of ANOVA F test A. Growth A.1. Sales growth A.2. Employment growth B. Participation in R&D programs as a percentage of sales B.1. Participation in domestic R&D programs as a percentage of sales B.2. Participation in European R&D programs as a percentage of sales C. Input R&D C.1. QSEs employed over total employees C.2. Gross R&D investment as a percentage of sales D. Output R&D D.1. Number of patents D.2. Introduction of new products and services E. Cost per job E.1. Cost per job

10 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) Results show that there are significant differences in three of the five categories of performance used and in five of the nine incubator performance indicators chosen. Specifically, we found significant differences in the participation in R&D programs as a percentage of sales (F=7.121; p=0.000) and in the participation in European R&D programs as a percentage of sales (F=3.478; p=0.021). Furthermore, we found significant differences in the percentage of QSEs as a percentage of the total number of employees (F=2.777; p=0.048). Finally, significant differences can be appreciated in the percentage of gross R&D investment as a percentage of sales (F=5.704; p=0.002) and the introduction of new products and services (F=3.782; p=0.014). We did not find significant differences in the number of patents (F=0.414; p=0.743), sales growth (F=0.909; p=0.453), employment growth (F=0.247; p=0.863) or cost per job (F=0.682; p=0.567). Once differences were found, we applied the Tukey multiple comparison test to refine results and determine which types of incubator firms showed significant performance differences. Tables 6, 7, 8, 9 and 10 include the results of the Tukey analysis providing information of pair wise differences between the four types of incubator with respect to the five categories of performance measures used Growth performance Mean sales growth and employee growth across the four types were 8.83% and 4.31% respectively. However, no significant differences were found in sales or employee growth among firms in the four types of incubator. Growth in the number of employees offered the higher non-significant difference. Specifically, as shown in Table 6, the highest non-significant difference in employee growth was found between private and economic development incubators. This result is to be expected as private incubators tend to have the largest growth oriented resources due to parent company resources. Large companies focus on sales growth as a proxy for profitability. Also, when human resources are needed by private incubators the parent company would typically contribute in the form of further funds or transferring executives or technical personnel to the incubator firm to meet objectives. On the other hand, economic development incubators, as described above, focus more on the promotion of entrepreneurship in areas with below average economic indicators. It implies that incubated firms are weaker in terms of idea generation, innovation and entrepreneurial potential. Given the lower entrepreneurial potential, those firms offer lower employee growth. Also the less competitive nature of firms in economic development incubators produces a lack of resources, which triggers a lower employee acquisition. Employees are not hired ahead of growth but when there is a need as the firms' resources are scant. Additionally, a low average sales growth (2.9%) implies that employees are not needed at the same rate as firms in basic research incubators, in which average sales growth is higher (9.88%) Participation in R&D programs The largest significant differences of all performance measures were found in the participation in R&D programs as a percentage of sales, as shown in Table 7, finding the highest significant differences in the participation of domestic R&D programs. University incubator firms lead the four types of incubator firms as they are the top receivers of funds through their participation in domestic R&D programs. They receive significantly more funds than basic research incubators (p=0.000), economic Table 6 Signification in differences between performance measures by growth performance indicators. Independent variable (I) TYPE2 (J) TYPE2 Mean difference (I J) Std. error Sig. Sales growth Basic research University Ec./Reg. Dev Private University Basic research Ec./Reg. Dev Private Economic/regional development Basic research University Private Private Basic research University Ec./Reg. Dev Employment growth Basic research University Ec./Reg. Dev Private University Basic research Ec./Reg. Dev Private Economic/regional development Basic research University Private Private Basic research University Ec./Reg. Dev

11 898 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) Table 7 Signification in differences between performance measures by participation in R&D Programs as a percentage of sales performance indicators. Independent variable (I) TYPE2 (J) TYPE2 Mean difference (I J) Std. error Sig. Participation in domestic R&D programs as a percentage of sales Participation in European R&D programs as a percentage of sales Basic research University Ec./Reg. Dev Private University Basic research Ec./Reg. Dev Private Economic/regional development Basic research University Private Private Basic research University Ec./Reg. Dev Basic research University Ec./Reg. Dev Private University Basic research Ec./Reg. Dev Private Economic/regional development Basic research University Private Private Basic research University Ec./Reg. Dev development incubators (p=0.014) and private incubators (p=0.011). University incubator firms receive the help of Technology Transfer Offices. The function of these institutions is, among others, to provide firms with information on funding and subsidies and they are extremely active on this front. Technology Transfer Offices are resource intense (information, personnel, etc.) and mostly receive public funding. Technology Transfer Offices are coordinated by the regional government to keep up-to-date on subsidies. Firms receiving the lowest percentage of funds through R&D participation are those located in private incubators. In this category we observe that some firms receive substantial funding but most do not receive any funding. The application for funding in R&D programs is a slow process which requires the commitment of resources which can end up with a less than certain result. Private incubators operate better in private capital markets. They can count on another type of funding (parent company capital, venture capital, etc.) which makes them less likely to receive funding of a public nature. Aside from private incubators, basic research incubators participate the least in domestic R&D programs (2.93%). Access to R&D program funds requires technical skills which by definition basic research incubators lack. Table 8 Signification in differences between performance measures by input R&D performance indicators. Independent variable (I) TYPE2 (J) TYPE2 Mean difference (I J) Std. error Sig. QSEs employed as a percentage of total employees Gross R&D investment as a percentage of sales Basic research University Ec./Reg. Dev Private University Basic research Ec./Reg. Dev Private Economic/regional development Basic research University Private Private Basic research University Ec./Reg. Dev Basic research University Ec./Reg. Dev Private University Basic research Ec./Reg. Dev Private Economic/regional development Basic research University Private Private Basic research University Ec./Reg. Dev

12 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) Table 9 Signification in differences between performance measures by output R&D performance indicators. Independent variable (I) TYPE2 (J) TYPE2 Mean difference (I J) Std. error Sig. Number of patents Basic research University Ec./Reg. Dev Private University Basic research Ec./Reg. Dev Private Economic/regional development Basic research University Private Private Basic research University Ec./Reg. Dev Introduction of new products and services Basic research University Ec./Reg. Dev Private University Basic research Ec./Reg. Dev Private Economic/regional development Basic research University Private Private Basic research University Ec./Reg. Dev In terms of the participation in European programs as a percentage of sales, again it is university incubator firms which show significant differences with respect to basic research incubators (p=0.026) and economic development incubators (p=0.060). Such a difference is more marked in European Union programs because firms need to be more informed and more competitive in order to receive funds from these types of programs. Information on European programs is more difficult to access than information on domestic programs. The link university incubators have with regional government institutions makes it easier to help firms succeed in the participation in European R&D programs. Additionally, European program calls are more competitive than domestic programs, as firms compete with firms from all over the European Union. University Technology Transfer Offices help university incubator firms by providing information on aspects such as which programs are available at European level, how to structure proposals to receive funding and through higher level networking they have access to information on which firms from other European Union countries are interested in jointly participating on a call proposal. This type of information makes university incubator firms more competitive when accessing funds from European R&D programs. Our research shows that participation in European R&D programs is non-existent for economic development incubator firms. Participation in European R&D programs requires even more technical skills than participation in domestic R&D programs as calls are more competitive. The risk of participation is very high also as scant resources owned should be contributed to a low certainty process Input R&D Two measures are used to assess performance on input R&D: QSEs as a percentage of total employees and Gross R&D investment as a percentage of sales. Table 8 shows that differences in the QSEs as a percentage of total employees are slightly significant for firms in private incubators when compared to basic research (p=0.072) and economic development firms (p=0.095). Table 10 Signification in differences between performance measures by cost per job performance indicator. Independent variable (I) TYPE2 (J) TYPE2 Mean difference (I J) Std. error Sig. Cost per job Basic research University Ec./Reg. Dev Private University Basic research Ec./Reg. Dev Private Economic/regional development Basic research University Private Private Basic research University Ec./Reg. Dev

13 900 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) Private incubator firms employ a higher number of QSEs compared with total employees and economic development employ the least. These results come as no surprise because we would expect economic development incubator firms to employ the least QSEs as a percentage of total employees. This type of incubator hosts firms created more by a survival driver than by an opportunity driver. Ideas are less innovative and belong to mature sectors. Hence the personnel are less skilled than other types of incubator described. Private incubators offer the highest percentage of QSEs. Private incubators count on resources to hire top talent and staff retention policies to keep them once they have joined. Additionally, private incubator firms can source top technical personnel recruited from internal capital markets. Differences are striking compared with economic development incubators. No significant differences were found in QSEs as a percentage of total employees between private incubator firms and university or basic research incubator firms. The mean QSE of total employees is lower, however, for these last two types. We could expect both, however, to rank higher in this performance indicator. University incubator firms can easily reach a large pool of QSEs within the university. Basic research incubator firms concentrate on high-tech and intense IP industries which are characterized by high percentages of highly skilled employees. Higher significant differences are found in the gross R&D investment as a percentage of total sales. Using this performance measure, university incubator firms rate higher than basic research (p=0.012), economic development (0.000) and private incubators (p=0.020). As university incubator firms are top participants in domestic and European R&D programs, they invest heavily in R&D. Mean R&D investment of these types of firms is 69.54% of sales. Basic research and private incubator firms are also heavy investors in R&D, as they invest 35.03% and 24.56%. Private incubator firms are more balanced as parent company management is aware of the importance of investment which yields high growth rates. R&D investment is lower than other incubator firms but efficiency is greater as they are extremely focused on the commercial side. Economic development incubator firms invest the least of the four types as they only invest a gross R&D 1.75% of sales Output R&D No significant differences were found in the number of patents as all incubator firms produce similar mean patents (see Table 9). Economic development incubator firms produce the least number of patents (mean=0.73) and basic research incubator firms produce the most (mean=4.00). University and private incubators also rank high in the production of patents with a mean of 3.82 and 2.22 respectively. It is worth noting that in meetings with incubator managers, they described how patent seeking was being substituted for speed to market, as companies believe the protection of intellectual property rights is a challenge. We appreciated how innovation agents lacked faith in the existence of a strong IP protection legal system. Finally, regarding the introduction of new products and services, basic research incubator firms showed significant differences with respect to university incubator firms (p=0.063) and to economic development firms (p=0.04). By definition, economic development firms develop less intense innovation processes than the other three types of incubator firms. Naturally, basic research incubator firms are more likely to introduce new products and services, followed by private incubator firms. In the latter case, pressure from the parent company yields greater new product and service launch capabilities Employment generation cost We did not find significant differences in the employment generation cost among incubator archetypes: The results of the ANOVA test for the cost per job did not show significance (F=0.682; p=0.567). Table 10 provides the results of the Tukey analysis determining differences among the four types of incubators. The average cost per job for the incubators in our sample was The figure is within the range of other European works. The European Commission Enterprise Directorate General provides a range of between 4000 and 8000 [45]. The European Business and Innovation Centre Network estimates a higher average cost for 2008 of 10,839 [48]. Research studies tend to show cost per job in the US as lower than in Europe. In the US, Markley and McNamara [44] estimated cost per job at $6580, while DiGiovanna and Lewis [47] estimated $3000. We agree with Monkman [49] that the cost per job in US incubators is lower than that of European incubators. Campbell and Allen [43] showed how cost per job may vary based on the location of the incubator. We do not expect the region to bias our sample as all the incubators in our sample operate in the same region. As shown in Table 4, average costs per job for basic research, university, economic development and private are 5537, 5371, 4087 and These results are logical; basic research and economic development incubators invest more heavily in R&D, as they receive quantum subsidies from different sources which lead to high costs. However, given the nature of the products and services they generate, technology related, they are likely to have a low cost per job in the long run. Technology generates high economies of scale and technology intense incubators are likely to benefit from such in the long run. Private incubators cost per job are the lowest. As explained above, these incubators are related to corporate development programs of larger companies. The lower cost per job derives from permanent cost consciousness and lean operation present in larger companies. 5. Discussion: Objectives versus performance Assessing incubator performance results in a need to distinguish between the different archetypal otherwise we would be comparing apple to pears [4]. We cannot conclude that incubator performance is satisfactory when different archetypes of incubators are included in the sample. The need to distinguish the archetype is necessary.

14 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) How should we proceed to determine whether each archetypal performance is appropriate? Performance should be compared to archetypal objectives. Objectives are very important in incubator literature [6,8,42] being the most widely used, and sometimes the most important criteria [4] to distinguish between archetypal business incubators. Archetypal success is the type which meets the objectives for which it was created. Next we assess performance by comparing incubator objectives to performance measures. As discussed in our Spanish incubator archetypal section, the basic research incubators' main objective is to develop technology, therefore, the archetype shows heavy investment in R&D as a percentage of sales. This archetype is also the one achieving a greater number of patents and launching more new products and services. So funds are well invested and investment yield is excellent when compared to other archetypes. Basic research archetype is not very aggressive in capturing R&D funds, however, it does not seem to have an impact on performance. Other archetypes, like university incubators capture more funds through the participation in European or domestic R&D programs, as they count on considerable aid through the University Technology Transfer Offices. University incubators secure more European or domestic funds of all the archetypal incubators. They are also very good at securing personnel as the largest percentage holder of QSE. However, this excess of resources is not reflected in the archetype obtaining the greatest number of patents or new products. In terms of patents, they rank second. However, in terms of launching new products they rank third. We could say that the university incubator objective is satisfactorily met as such objective is the licensing of IP researched within the university. The objective of spinning off seems to lag as the incubator does not lead in the launch of new products. Based on the above argumentation, both archetypes, basic research and university incubators, meet the objectives pursued. However basic research incubators successfully meet the objectives whereas university incubators only moderately meet the objectives. Economic development incubators clearly do not meet objectives, as performance is lowest in every category. Incubator objectives of regional development through the creation of wealth and employment can be best assessed in terms of sales growth and employment growth. Economic development incubators rank third in the generation of employment, only followed by the university archetype whose main objective is not the generation of employment. The economic development incubator does not achieve wealth creation either as the archetype develops ventures with the slowest sales growth. In absolute terms, average sales and employment growth of 2.99% and 2.37% are substandard when compared to regional GDP growth or inflation numbers for the period of study (3.2% and 3.3% respectively). Finally, private incubators do meet set objectives as they create high returns for the parent company. This can be derived from growth, the launch of new products and patent generation figures. Sales growth is a proxy for venture success [30,31] and sales growth rates in the region of 20% for a five year period can be deemed as reasonably successful. Launch of new products leads to higher profitability. Patent generation is a proxy for competitive advantage. Private incubators rank first in sales growth, second in new product launch and third in patent generation. In summary, from our research we can conclude which archetypal incubators perform best by meeting the objectives for which they were created. The least successful archetype is economic development as objectives of employment and wealth creation are clearly not met. Basic research and private incubators satisfactorily meet objectives pursued. University incubators moderately meet their targets. 6. Conclusions Our research leads us to conclude that not all archetypes perform equally but that there are significant differences in the performance of the different archetypes. Some types perform better in specific performance measures while others perform worse. Our research leads us to conclude that not all incubators perform satisfactorily. Performance depends on the archetypes' objectives and some archetypes achieve the objectives for which they were set up and other archetypes do not. Economic development incubator performance is substandard. University incubators perform satisfactorily. Private incubator and basic research incubators performance is outstanding. 7. Future research Public policy and research cannot ignore the effects produced by the heterogeneity of an incubator typology. Further research is needed in determining when each type of incubator should be used. In order to refine incubator policy a mix of incubator types should be considered thoroughly. Before getting to a formula leading to an improved combination of incubators, more studies unearthing the differences between characteristics, performance and unique features of incubator types should be revealed. Some important public policy questions on incubation are: how many incubators should a country/region open and where should they be placed?; What is the optimum mix of private/public incubator is optimal?; and lastly, what proportion of each type would lead to efficient public/policy decisions? Allen and McCluskey [5] mention incubation strategy as part of the incubator manager job description. We believe this belongs to a higher hierarchical position. But before officials can improve their decision making process, we need to provide them with more empirical evidence on the features and performance of incubator types. References [1] P. Westhead, R&D inputs and outputs of technology based firms in science parks, R&D Manage. 27 (1) (1997) [2] S.A. Mian, Assessing and managing the technology business incubator: an integrative framework, J. Bus. Venturing 12 (4) (1997) [3] M.T. Hansen, H.W. Chesbrough, N. Nohria, D.N. Sull, Networked incubators. Hothouses of the new economy, Harv. Bus. Rev. 78 (5) (2000) [4] R. Aernoudt, Incubators: tool for entrepreneurship? Small Bus. Econ. 23 (2) (2004) [5] D. Allen, R. McCluskey, Structure, policy, services and performance in the business incubator industry, Entrepreneurship Theory Pract. 15 (2) (1990) [6] R. Grimaldi, A. Grandi, Business incubators and new venture creation: an assessment of incubating models, Technovation 25 (2) (2005) [7] B. Becker, O. Gassmann, Gaining leverage effects from knowledge modes within corporate incubators, R&D Manage. 36 (1) (2006) 1 16.

15 902 J.L. Barbero et al. / Technological Forecasting & Social Change 79 (2012) [8] E. Carayannis, Von Zedtwitz, Architecting glocal (global local), real-virtual incubator networks (G-RVINs) as catalysts and accelerators of entrepreneurship in transitioning and developing economies: lessons learned and best practices from current development and business incubation practices, Technovation 25 (2) (2005) [9] M. Colombo, M. Delmastro, How effective are technology incubators? Evidence from Italy, Res. Policy 31 (7) (2002) [10] D. Siegel, P. Westhead, M. Wright, Science Parks and the performance of new technology based firms: a review of recent UK evidence and an agenda for future research, Small Bus. Econ. 20 (2) (2003) [11] A. Bollingtoft, J. Ulhoi, The networked business incubator leveraging entrepreneurial agency? J. Bus. Venturing 20 (2) (2005) [12] M. Schwartz, C. Hornych, Specialization as strategy for business incubators: an assessment of the Central German Multimedia Center, Technovation 28 (7) (2008) [13] K. Aerts, P. Matthyssens, K. Vandenbempt, Critical role and screening practices of European business incubators, Technovation 27 (5) (2007) [14] P. Westhead, S. Batstone, Perceived benefits of a managed science park location for independent technology-based firms, Entrepreneurship Reg. Dev. 11 (2) (1999) [15] M. Von Zedwitz, R. Grimaldi, Are service profiles incubator-specific? Results from an empirical investigation in Italy, J. Technol. Transf. 31 (4)(2006) [16] B. Clarysse, M. Wright, A. Lockett, E. Van de Velde, A. Vohora, Spinning out new venture: a typology of incubation strategies from European Research Institution, J. Bus. Venturing 20 (2) (2005) [17] A. Bergek, C. Norrman, Incubator best practice: a framework, Technovation 28 (1 2) (2008) [18] S.A. Mian, Assessing value added contributions of university technology business incubators to tenant firms, Res. Policy 25 (3) (1996) [19] M.J. Nowak, C.E. Grantham, The virtual incubator: managing human capital in the software industry, Res. Policy 29 (2) (2000) [20] M. Rice, Coproduction of business assistance in business incubators. An exploratory study, J. Bus. Venturing 17 (2) (2002) [21] P. Phan, D. Siegel, M. Wright, Science Parks and incubators: observations, synthesis and future research, J. Bus. Venturing 20 (2) (2005) [22] I. Peña, Business incubation centers and new firm growth in the Basque country, Small Bus. Econ. 22 (3 4) (2004) [23] B. Bigliardi, I. Dormio, A. Nosella, G. Petroni, Assessing Parks performances: directions from selected Italian case studies, Technovation 26 (4)(2006) [24] J.A. Hansen, Innovation, firm size, and firm age, Small Bus. Econ. 4 (1) (1992) [25] A. Markusen, P. Hall, A. Glasmeier, High Tech America: The What, How, Where, and Why of the Sunrise Industries, George Allen and Unwin, Boston, USA,1986. [26] A.K. Chakrabati, Scientific output of small and medium size firms in high-tech industries, IEEE Trans. Eng. Manage. 37 (1) (1990) [27] M.E. Mogee, R.G. Kolar, International patent analysis as a tool for corporate technology analysis and planning, Technol. Anal. Strateg. Manage. 6 (4) (1994) [28] F.T. Rothaermel, M. Turnsby, Incubator firm failure or graduation? The role of university linkages, Res. Policy 34 (7) (2005) [29] P. Westhead, D.J. Storey, An Assessment of Firms Located on and Off Science Parks in the United Kingdom, HMSO, London, UK, [30] F. Delmar, Measuring growth: methodological considerations and empirical results, in: R. Donckels, A. Miettinen (Eds.), Entrepreneurship and SME Research: On its Way to the Next Millennium, 1997, pp [31] F. Delmar, P. Davidsson, W.B. Gartner, Arriving at the high-growth firm, J. Bus. Venturing 18 (2) (2003) [32] F.T. Rothaermel, M. Turnsby, University incubator firm knowledge flows: assessing their impact on incubator firm performance, Res. Policy 34 (3) (2005) [33] M. Marinazzo, Science park evaluation and organizational analysis, The Science Park Evaluation Handbook, European Innovation Monitoring System (EIMS), Publication no. 61,, [34] S. Radosevic, M. Myrzakhmet, Between vision and reality: promoting innovation through technoparks in an emerging economy, Technovation 29 (10) (2009) [35] D.J. Storey, B.S. Tether, Public policy measures to support new technology-based firms in the European Union, Res. Policy 26 (9) (1998) [36] M. Porter, On Competition, Harvard Business School Press, Cambridge MA, USA, [37] A.N. Link, J.T. Scott, Science parks and the generation of university-based knowledge: an exploratory study, Int. J. Ind. Organization 21 (9) (2003) [38] H. Löfsten, P. Lindelöf, Science parks and the growth of new technology-based firms academic industry links, innovation and markets, Res. Policy 31 (6) (2002) ; H. Löfsten, P. Lindelöf, Growth, management and financing of new-technology-based firms assessing value-added contributions of firms located on and off science parks, Omega 30 (3) (2002) [39] P. Dettwiler, P. Lindelöf, H. Löfsten, A comparative survey between small new technology-based firms located on and off Science Parks implications for facilities management, Technovation 26 (4) (2006) [40] P. Escorsa, J. Valls, Tecnologia i innovació a l'empresa, Direcció i gestió, Edicions UPC, Barcelona, Spain, [41] R.W. Smilor, M.D. Gill, The new Business Incubator: Linking Talent, Technology, Capital, and Know-how, Lexington Books (Lexington, Mass), USA, [42] L. Aaboen, Explaining incubators using firm analogy, Technovation 29 (10) (2009) [43] C. Campbell, D. Allen, The small business incubator industry: microlevel economic development, Econ. Dev. Q. 1 (2) (1987) [44] D.M. Markley, K.T. McNamara, Economic and fiscal impacts of a business incubator, Econ. Dev. Q. 9 (3) (1995) [45] European Commission Enterprise Directorate General, Benchmarking of business incubators: final report, 2002 (Brussels). [46] U.S. Department of Commerce Economic Development Administration, Construction grants program impact assessment report, [47] S. DiGiovanna, D.A. Lewis, The Future of Technology Incubation in New Jersey: A Strategy for the New Jersey Commission on Science and Technology, Project on Regional and Industrial Economics, Rutgers University, New Brunswick, NJ, [48] European Business and Innovation Centre Network, The BIC network in 2008: Facts and figures, Brussels, [49] D. Monkman, Business Incubators and their Role in Job Creation, NBIA publications, OH, Dr. Jose L. Barbero is an associate professor at Universidad Pablo de Olavide (Seville). He has previously been professor chair of entrepreneurship at EOI Business School. He holds a degree in law and masters in Business Administration (ESADE) and finance (London Business School). He has received a Fulbright scholarship to study at Harvard University. Dr. Barbero is the author of two books and has served as a Visiting Scholar and Researcher at New York University (NYU). He is currently doing research in the area of incubation, venture growth, internationalization and corporate turnaround. Dr. Jose Carlos Casillas is a professor at the University of Seville (Spain). He is co-chair of family business at the University of Seville. Jose Carlos is an economist and has researched extensively on entrepreneurship and published in the top entrepreneurship journal. His research interests are in the area of venture growth, incubation, knowledge management and internationalization. Alicia Ramos is a Doctoral student at the University of Seville. She has worked in the finance area in Spain and the United Kingdom. She earned a bachelor in business administration at the University of Seville and has studied an advance program in finance in London Business School. Her dissertation relates to types of incubators and types of entrepreneurs. Susana Guitar is the Director of Innovation at the Andalusian Regional Government in Spain. She is a telecommunications engineer and holds an MBA by Instituto de Empresa. She has previously been director of environmental programs at EOI Andalucia Business School and has held management positions at diverse companies.