Too big to fail? A case study of the rise and fall of a medical research infrastructure

Sammanfattning: Introduction: Distributed Research Infrastructures are becoming increasingly more salient as science expands and universities continue to look for new means to cooperate and share expertise and expenses on large-scale projects. One area which has seen much development in recent years is biobanking, as there have been numerous attempts to harmonise the different biobanking standards over the years, none of which have been entirely successful. BBMRI.se was an EU-initiative that sought to harmonise the biobanks nationwide. BBMRI.se, was thus selected as a case for studying how a distributed Research Infrastructure was set up. At the time of its creation, the organisation constituted the largest investment ever made by the Swedish Research Council in a medical Research Infrastructure. The organisation involved all Swedish universities with a medical faculty, in addition to two other universities. However, the organisation was marred by a number of controversies and would eventually fold in 2018. Aim: This dissertation is to elucidate the mechanisms involved in the construction of a medical large-scale distributed Research Infrastructure, and to understand the motivations and rationale of the experts who activate themselves in constructing it. Thus, the overall aim of this doctoral thesis is to identify the benefits and constraints of forming a large-scale medical, distributed Research Infrastructure. Specifically, this dissertation looks at a real-life case while comparing it to the available literature covering the development of Research Infrastructures as well as some of the theories covering mindsharing and collective entrepreneurship. The ambition is to contribute knowledge on the determining factors in bringing a large-scale infrastructure together as well as the risks associated with it. Hence, this dissertation asks the following research question: What are the principal lessons for researchers, entrepreneurs and funders that can be inferred from the formation of a large-scale distributed Research Infrastructure towards securing more sustainable prospects for similar, future endeavours? More precisely, this dissertation seeks to determine what the most debated topics are within the academic discourse on Research Infrastructures (study I), after which it looks at the factors involved in constructing shaping a distributed Research Infrastructure (study II). The study then endeavours to looks at some of the pitfalls and how managerial self-governance affects organisational failure (study III). The study then seeks to investigate the mind-set of the managers/pioneers involved in setting up BBMRI.se and if they perceive the organisation in a similar fashion for the other managers (study IV) and how they have reasoned behind their motivations for joining the initiative in the first place (study V). The overall results have endeavoured to elucidate what components are at work when forming such an infrastructure at an organisational level, but also to understand the reasoning and motivation that the individuals responsible in setting up the infrastructure might have had, and how their visions and/or actions may have impacted on the organisation. Method: Some various designs and data collection methods were used in this dissertation. Study I was a literature study carried out as a narrative review using the PRISMA statements as a guideline. Both the Web of Science (WOS) and PubMed databases were scoured for articles. Study II-V used qualitative, semi-structured interviews with BBMRI.se managers. All of these studies took on the form of iterative, directed content analyses, with the exception of study III, which was an inductive, directed content analysis. Results: Study I found that the most commonly discussed topics concerned the need for developing and expanding the use of “infrastructures”. The findings indicated that the future of scientific research calls for a deeper and more widespread multidisciplinary forms of collaboration. Study II found that it is crucial to identify the potential collaborative and deliberative organisational elements of organisational team building already at the outset of establishing a distributed Research Infrastructure. The study also found that, contrary to suggestions of extant literature, the establishment of a distributed Research Infrastructure does not necessarily counteract organisational fragmentation. Study III identified that an organisation with high levels of task uncertainty and low levels of organisational integration will suffer from organisational fragmentation. The type of fragmentation manifested in BBMRI.se is best identified as a “fragmented adhocracy”. This means that the organisation’s mission statement is subject to diverse views, leading to goals that are separate, unstable and sometimes even conflicting, while also lacking in co-ordination. The study also found that the organisation lacked a “liaison device” and instead depended on a more traditional model of planning and control systems through its reliance on strategy documents and interim evaluation reports. This was in spite of the fact that this model is better suited for a more vertical organisational structure. Study IV investigated how managers/associates of BBMRI.se perceived the organisation’s brand and the role of “mindsharing”. The results showed that mindsharing occurred throughout the initial two stages (“Brand Strategic Analysis” and “Brand Identity”), but would disspate throughout the remaining two final stages (“Brand Operationalising”, and “Post Implementation Reflections”). This resulted in a fragmented brand perception, which resulted in the failure of generating a “pull-effect” for the BBMRI.se brand. Study V looked at how collective entrepreneurial team cognition of the instigators behind BBMRI.se changes throughout the process of establishing the organisation. The study devised a new “action phase model”, known as the “4 I’s” of entrepreneurship, where each “I” elaborated on the entrepreneurial rationale behind the various stages of the creation process. These were “Intention”, “Initiation”, “Implementation” and “Introspection”. The results illustrated that the respondents agreed that there was a need for BBMRI.se, while disagreeing on what the organisation should be doing and what its challenges consisted of. The homogenous mind-set would begin to dissipate once the “Initiation” stage was reached, declining further throughout the Implementation stage. Conclusion: The overall conclusions from study I-V have shown that distributed Research Infrastructures carries potential to form a platform to pool scientific research in the face of the ever-expanding sciences, where the demands of co-financing and scientific co-operation are becoming ever so pressing. In addition, distributed Research Infrastructures have the benefit of utilising initial synergy effects and using multidisciplinary teams. In line with the contention of Gibbons et al. (1994), this carries the potential of opening up new possibilities of scientific knowledge production. Provided there is a political incentive in place to allocate the necessary funding, the process of establishing a distributed Research Infrastructure can be done in a considerably swift timespan. However, there are several inherent risks. Most notably, there was a lack of “infrastructuring”, as defined by Star and Bowker (2002). This means that scientists as well as the policy- makers should gradually learn together through a learning process about how to creating an effective large-scale infrastructure. This may have prevented mindsharing from becoming consolidated throughout the formation process (Aaker, 1996; Acuña, 2012; Azevedo, 2005; J. Griffin, 2009; Holt, 2016; Krishnan, Sullivan, Groza, & Aurand, 2013; Stevens, 2003). This, in turn, would also put an end to the collective entrepreneurship that had up till that point characterised BBMRI.se, in which the motivations and drivers of the initiators/managers, as well as their respective recollections of the same, were instrumental features (Cardon, Post, & Forster, 2017; Czarniawska-Joerges & Wolff, 1991; Sakhdari, 2016). Moreover, the integration of autonomous “National Champions” (leading scientists within their field) carries a risk of the “principal-agent” problem, which in turn can lead to “moral hazard” as the “National Champion(s)” may elect to undertake added risks, since someone else bears the cost of those risks (Holmstrom, 1982; Laffont & Martimort, 2002; Steets, 2010). There is also an over- whelming risk of organisational fragmentation, which, coupled with managerial neglect, may cause the eventual failure of the organisation.