THE BIOSCIENCES KNOWLEDGE VALUE CHAIN AND
COMPARATIVE INCUBATION MODELS
Authors: Phil Cooke*, Dan Kaufmann**, Chen Levin**
*Centre for Advanced Studies **Jerusalem Institute for
Cardiff University Israel Studies
Cardiff CF10 3BB Jerusalem 92 186
March 2003
Prepared for Regional Studies Association Conference on ‘ Reinventing Regions in the Global Economy’ Scuola Superiore Sant’ Anna, Pisa, Italy, 12-15 April
1. Introduction
1.1 This research derives from an EU DG Enterprise (IPS Programme) project on bio-incubation, called Bio-Link. The Bio-Link project is innovative in three ways. First it involves an international comparative analysis of biotechnology incubators of the kind that is rarely if ever done. Second the incubator representatives are monitored and investigated by an academic partnership team. Third, there is a stated aspiration by the incubator companies to engage in co-incubation across borders. Co-incubation is, as far as we are aware, a new kind of boundary crossing innovation in which advanced start-up businesses are assisted to enter other national markets and/or benefit from specialised services or scientific, technological, or commercial knowledge absent in the home country but present in a partner country.
1.2 This paper summarises the first stage of the project conducted during the first three months of 2003. This involved detailed presentation, monitoring, analysis and evaluation of incubators by type, by service offered, by ownership and by position in what can be called the knowledge value chain for Life Sciences more generally, and biotechnology more particularly. A pilot conceptual model of such a ‘knowledge value chain’ is presented in the next section. It passes from exploration through examination to exploitation knowledge as basic research evolves to the stage of Pre-Clinical and Clinical Trials and then to the Commercialisation stage where due diligence skills and venture capital skills ultimately determine whether or not a ‘prospected’ start-up may reach the stock market, license its technology or be acquired by another owner.
1.3 Following this expository section, the paper then moves into the main stage of comparing and contrasting, monitoring and evaluating the five incubators in the study, recording the judgements of both academic observers and incubator managers regarding their specific incubator type, strengths and challenges, or preferred market position. An attempt is made to resolve these judgements where variation occurs. There is always an attempt both to situate the bio-incubators in the wider local or regional innovation system within which they reside, and to recognise that where one may offer a wider range of services to academic entrepreneurs, this may be a sign of system weakness rather than strength, although for other reasons – ‘the one-stop shop’ – it may of course be a sign of efficient focusing of many support functions under one roof.
1.4 The UK is said to be Europe’s closest equivalent to North America in business climate, but recent reforms in Israel, including a swift growth in venture capital and a sustained programme of investment in incubation facilities led by the government, particularly advised by the Chief Scientific Officer (CSO) may have placed Israel even closer to the entrepreneurial climate of North America. Contrariwise, Israel, as a small economy, lacks some key ‘knowledge value chain’ components like pharmaceuticals firms, Clinical Research Organisations (CROs for Clinical Trials), and biomanufacturing facilities. Yet, for incubation facilities, Israel is relatively well-provided.
1.5 Some efforts at international comparison on key indicators comprises a brief fourth section of the report. A conclusion sums up the position discovered thus far and points to the next project steps, notably a survey of incubator firms to hear their assessment of the bioincubation experience as they see it and their views on co-incubation.
2. A Model of the Knowledge Value Chain in Life Sciences and the Role of Incubation
2.1 The point of co-location practices by innovative start-ups and their support organisations, such as incubators, is to enhance the competitiveness of firms by increasing productivity and innovation. Proximity yields knowledge spillovers (Audretsch & Feldman, 1996; Feldman & Audretsch, 1999). Some of these do not travel very far, Jaffe et al. (1993) suggesting fifty miles for patent citations by entrepreneurs referring to university research. This has been confirmed for Europe by Senker & van Zwanenberg (2001). However, in addition to strong local spillovers in relations among researchers and entrepreneurs, US researchers enjoy more extensive network relationships among public research organisations than their European counterparts (Owen-Smith et al, 2000). But, either way, at least for biotechnology, the complementarities are conditioned very strongly by arm’s length exchange when valuable exploration knowledge is being transformed by exploitation knowledge as ‘star’ scientists interact on serious commercial projects, for example joint patenting or product development with biotechnology entrepreneurs (Zucker et al., 1998). This corrects somewhat for an ‘over-socialised’ view of the importance of
Fig. 1: Conceptual Model of the Bioscientific & Biotechnological Value Chain
Exploration
Knowledge
Examination
Knowledge
Exploitation
Knowledge
‘untraded interdependencies’ in cluster-type settings. This is not to deny knowledge spillovers that enable ‘free-riding’ happen, rather that they are relatively trivial in nature.
2.2 Nevertheless, for the acquisition of localised spillovers of a trivial or a non-trivial kind, a number of key knowledge spillover ‘types’ need to be available, and, following Porter (1998) and others, they are better if local because of time economies. These refer to:
· Swift receipt of value-adding knowledge prior to its general release (anticipatory knowledge)
· Timely availability of complementary local assets or capabilities (participatory knowledge)
· Early access to local inventions, discoveries or innovations (precipitatory knowledge)
These correlate to three more thoroughgoing kinds of knowledge central to knowledge production in research and commercialisation activity that are not spillovers but core knowledge production activities. These are shown down the left side of Fig. 1 and explained beneath it:
· Exploration knowledge – the aim of fundamental research, as, for example, in post-genomics, proteomics and molecunomics conducted in laboratories of universities, research institutes and DBFs in the main.
· Examination knowledge – the kind of ‘feedback’ knowledge that comes from clinical trials of new treatments, therapies and drugs to find out if they work or work better than existing treatments
· Exploitation knowledge – the mix of knowledge skills, including scientific, technological, entrepreneurial, financial and legal that enables discoveries to be transformed into commercial products with market demand.
2.3 Where clusters have the fullest range of these ‘knowledge production’ and ‘knowledge spillover’ capabilities they exert synergies upon a diverse range of scientific research and commercialisation actors that, on the one hand, causes further and further ‘swarming’ (à la Schumpeter) at the cluster epicentre and, on the other, boundary extension through network-like market relations to specialised support services in demand. These are externalised services such as biomanufacturing, bioimaging or clinical research organisation (CRO) services that function effectively at greater distance from the cluster epicentre. Indeed these may be points hitherto related to different industries and earlier economic histories, as with synchrotrons (particle accelerator/colliders), formerly key to nuclear physics but now more important for bioimaging, or biomanufacturing that has a lineage in bulk chemicals production. In Fig. 1 a conceptual model of such links according to knowledge spillover and production typologies is presented. That is, the three ‘E’s’ structure the vertical dimension of the knowledge value chain, from exploration, through examination, to exploitation process-phases. The three ‘I’s’ are embedded in the process linkages among actors in which Intelligence, Insurance and Investment appear as key rationales for proximity relations. These facilitate anticipatory, participatory and precipitatory spillovers respectively, as schematised in Table 1.
2.4 While a diagram should be worth a thousand words, it remains necessary to say
Table 1: Varieties of Knowledge Spillover in Bioscientific Value Chains
something about the content schematised above. We have seen, rather simply, how by escaping the strictures of the ‘tacit’ in relation to ‘codified’ knowledge distinction, we can gain more content regarding the nature of knowledge ‘management’ advantages accruing For example, econometric analysis of limited aspects of the schema reveal consistently
and across different national innovation systems, high associations between R&D expenditure and patenting activity (Jaffe, 1989; Audretsch, 2001; Fritsch, 2001; Andersson & Ejermo, 2002). There seems to be, accordingly, an R&D production function that is strongly geographically structured. This is explained in terms of the schema’s conceptual categories of ‘local anticipatory spillovers’ like ‘foresight’ or ‘a glimpse into the future’ arising from in situ exploration knowledge locally regulated or constituted by ethics (including trust and reputation, as well as ‘due research diligence’) and acted upon ex ante in the peer review decision to fund on grounds of research excellence (at proposal stage) and cognitive validity (successful) defence of patent application.
2.5 Elsewhere, less econometrics have yet been performed, so we must rely on stylised facts and textured case analyses. Thus examination knowledge is extremely under-researched by social scientists though it is a crucial part of the process of determining the reliability and validity of applied exploratory knowledge in the form of functional treatments (in biopharmaceuticals and pharmaceuticals, more generally). Some key cases indicate that while ‘big pharma’ still arranges pre-clinical and clinical trialling, DBFs are in stronger negotiating positions than hitherto and can insist on taking responsibility for this. Second, as noted, Clinical Research Organisations are increasingly important as outsourcing organisers of clinical trials as well as numerous other contract services for biopharmaceuticals firms. Though some, like Quintiles locate in clusters, as in Cambridge/Boston in the US and Edinburgh in Scotland (UK), others, like another UK firm Cobra are located in ‘no man’s land’ between Manchester and the Midlands of England, where there is no serious bioscientific critical mass.
2.6 Some trials require patients with very rare diseases or genetic combinations, so there is no proximity advantage from the presence of a large, general hospital nearby. However, in the Massachusetts case, the presence of the General Hospital in Boston, Brigham & Women’s and many others in the Greater Boston area is important to the cluster not only for patient trials but scientific partnership (‘Partner’s Healthcare’ was in 1999 a $400 million research programme involving these hospitals, Harvard Medical School and Genzyme, among others) and specific project research, especially for National Institutes of Health, an indication of the value of which is given in Table 2. Total 2000 NIH funding in Boston was $1.09 billion. Examples of research and exploitation partnerships between these and DBFs in the region include: Curis and Harvard University (genetic signalling); Genzyme and each of Massachusetts General Hospital (HIV/AIDS), Dana-Farber Cancer Institute and Beth Israel Medical Centre (melanoma clinical trials), and Ariad with the Whitehead Institute, Massachusetts Institute of Technology and Harvard University (Cell Sequencing research).
Rank Institution NIH Funding
7 Harvard University $250.4 million
17 Massachusetts General Hospital $180.5 million
22 Brigham & Women’s Hospital $162.5 million
38 Boston University $108.2 million
47 Dana-Farber Cancer Institute $87.2 million
53 Beth Israel Deaconess Medical Centre $82.1 million
54 Whitehead Institute for Biomedical Research $81.3 million
60 University of Massachusetts Medical School $73.9 million
58 Massachusetts Institute of Technology $75.0 million
74 Children’s Hospital $52.9 million
86 Tufts University $37.5 million
Table 2: Principal NIH-Funded Research Institutions in Massachusetts, 2000.
Source: National Institutes of Health
2.7 Finally, at the exploitation phase of the knowledge value chain, some econometric work has been conducted, notably by Zucker (1998) and colleagues.
They found the following regarding the propensity to cluster by DBFs and research scientists, notably those of ‘star’ status:
· Especially in the early years, the exploitation of biotechnology required the mastery of a very large amount of exploration knowledge that was largely non-codified. Thus firms (DBFs) became inordinately dependent on research scientists to ‘translate’ for them. The latter were well attuned to working with industry, hence receptive to such interaction. Locations with concentrations of such knowledge to transfer thus became magnets for DBFs as ‘big pharma’ an early user and facilitator of research discovered their own ‘absorptive capacity’ problems deriving from their origins in fine chemistry not biology
· ‘Untraded interdependencies’ or pure ‘knowledge spillovers’ (non-pecuniary) do not seem to apply in biotechnology. Discoveries do not transfer swiftly through social ties or informal seminars but rather display high ‘natural excludability’. This means biotechnology techniques are not widely known, so ‘stars’ exploit this by entering contracts with DBFs to exploit supranormal rates of return. Localisation arises as the scientist interacts with proximate DBFs because she usually retains affiliation to the academic home base.
· The innovative performance of DBFs is positively associated with the total number of articles by local university biotechnology ‘stars’. However, further data disaggregation of ‘stars’ into those contractually tied and untied to local firms show the positive association only applies to contractual collaborators, while the coefficient loses both significance and magnitude for the others.
2.8 Finally, regarding the precipitation spillovers dimension, that is the advantages of proximity for those firms that ‘make it happen’ i.e. turn a scientific finding into a firm that commercialises a drug, treatment or diagnostic test, namely venture capitalists, specialist lawyers and consultants, there is econometric and case study evidence that exploitation knowledge causes them to locate their investment a mean distance of one hour’s driving time from their office base for the most part. Thus Zook (2000) and Norton (2000) both showed distance decay effects setting in after that radius had been passed for high technology clustering (including biotechnology). Granovetter et al (2000) found very high network densities linking Silicon Valley firms and locally rather than nationally founded venture capitalists. Powell et al., (2002) confirmed that venture capitalists invest most in firms located nearby to maximise collaboration and knowledge gains from investments. Further research to measure the numbers and nature of linkages among venture capitalists in advanced clusters like Silicon Valley and small emergent clusters in less favoured regions like Northern Ireland by one of the present authors (Cooke, 2001; Cooke, Roper & Wylie, 2003) was also conducted. In both cases venture capitalists not only retain stock ownership in DBFs after IPO flotation, they actively encourage sub-contracting and a variety of forms of technology development, marketing and management linkages between DBFs in their portfolios. In the case of KPCB in California this keiretsu characteristic is elevated to corporate policy and website advertising. In less favoured places they do the same, unaware of its best practice nature, because it makes sense for the health of the DBFs and the venture capital company’s investment. The one-hour role predominates in both types of setting.