One of three very useful conclusions for those running or supporting incubators is: “there is fairly clear evidence incubators and accelerators work overall - for survival, employment growth and receiving external finance”.
But rather than skip to conclusions I recommend reading from the start. The opening paragraph alone is as valuable as the conclusions, as it succinctly summarises the arguments for innovation and entrepreneurship. And it continues, throughout the rest of the article there are many very useful summaries and clarifications including: co-location as complementary to clustering, effects on job growths, whether female and minority participants disproportionately benefit, multiple perspectives on the rise of coworking and evidence for benefits of geographic (or other forms of) "proximities".
The ‘What makes a good incubator?’ article seems a good complementary read, published around the same time and written without awareness of the CEP's work on the similar topic. The CEP discussion paper asks deeper though similar questions (“we need to consider who benefits”) and answers them with academic rigour with a city or region-scale economic development perspective.
I especially appreciated the description of the evidence of the benefits of proximities. Several sources are cited in concluding that “real world knowledge spillovers exhibit substantial distance decay, especially for ‘knowledge-intensive’ services…and spillovers may disappear within 250m”. And more than just geographical, ‘proximities’ include social closeness (e.g. through friendship), organisational (e.g. working in the same firm), cognitive (e.g. the same subject background), or institutional proximity (e.g. common norms).
This evidence around proximities makes sense of the 'feel' one gets about incubators that are working or aren't and the need for them to be 'buzzing', have 'density' or be 'centres of gravity'. This sense of ‘proximity’ also seems fundamental to the quality of ‘designing for and accelerating serendipity’ as often cited by coworking advocates and explored in this Future Tense podcast.
A second (of three) final conclusions in the article opens nicely into not knowing and perhaps the limits of what can be known through this sort of study and research: “we have much less clarity about how programmes achieve results”. This is aligned with the GALI studies comments about the impacts of different program design.
That this discussion paper is so thorough and excellent demonstrates how sound foundations enable next questions. For example: knowing that incubators work generally, we can now ask how and for whom do they work specifically?
I suspect questions like that lend themselves a bit more to ethnography, intense studies of the micro-details of 'customer journeys' and investigating the micro-dynamics and value created in individual and aggregated human interactions within incubation situations.
That question and investigation isn't just academic or of interest to public funders, but actually the daily concern of incubator and accelerator managers. It goes to the heart of what the best 'hosts' and incubator managers do: to in each interaction, feel for the alignment and contribution that will add value for the human in front of them, the venture, investor, incubator, city and community.