The emergence of modern innovation ecosystems
In my previous post, I introduced the Theory of Ecological Succession as one that applies just as well to innovation ecosystems as it does to natural ones. Just as scrapy pioneer species prepare the soil for more complex communities to develop in nature, so do scrapy entrepreneurs pave the way for intermediate institutions, and eventually complex innovation communities.
In the world of innovation, the most important anchor institution is the University. The University’s role as hub of innovation is a long and complex one that began evolving toward its current state after the second world war, when Vannevar Bush, then the Director of the US Office of Scientific Research and Development (superseded by the NSF) proposed that the then unprecedented government support of academic science be continued into the post-war era. The existential threat of the war years meant that academic scientists around the world were effectively conscripted to support their country’s military in developing technologies for war. Those expelled by their country of origin, including, decisively, most of continental Europe’s Jewish scientists, took up positions in the US or UK where, in all but a few cases, they were quickly integrated into the military-scientific machine. Never-before-seen sums of money were lavished on scientists and engineers. In return the government was to decide, at least at a high level, how it was spent.
Institutions that are now seen as the bedrock of innovation in the US and the world — MIT, Harvard, Stanford, UC Berkeley — solidified these positions by hosting massive military research programs on their campuses, in effect calling “dibs” on the best and brightest scientists and engineers available. MIT’s Rad Lab, Harvard’s Computation Lab, Berkeley’s Radiation Lab, and later, MIT’s Semi-Automatic Ground Environment air defense system (SAGE), employed hundreds of scientists each and became central to the inventions of RADAR, SONAR, nuclear energy, and computers, as well as a host of new materials and devices.
The end of war in 1945 meant the prospect of the government winding down support for research to pre-war levels. Universities across the country would be left with empty buildings and scientists of all stripes left without jobs. No doubt already hearing rumblings of this from his perch in Washington, DC, Vannevar Bush penned a now-famous letter to President Roosevelt arguing that government support for science should not be diminished after the war, but simply re-oriented toward fighting a new “war of science against disease.” Bush’s proposal, though not implemented in its details, laid the conceptual groundwork for the centrality of the NSF and NIH in postwar research. Both his letter, since published as “Science, the Endless Frontier”, and Roosevelt’s response merit a read by anyone interested in the history of scientific research. To this day, the majority of basic research research in the US is conducted by professors and graduate students on university campuses, funded in large part by government agencies. Roosevelt’s initial vision of a “war of science against disease” also persists; more than half of all government funding for research remains dedicated to advancing human health.
One other major government intervention is important to note as we consider how University-based innovation came to be what it is today: the Bayh-Dole act of 1980. When government agencies took center stage in funding university research post World War II, they also became some of the largest recipients of US patents; Anything invented using government funds was, by default, government property. This posed a problem, however, as the government was unwilling to grant exclusive patent rights to any company. How could you give private entities exclusive access to technologies owned and paid for by taxpayers? But without exclusive rights, private entities were unwilling to invest the large sums of money required to further develop and commercialize them. Why would they? Any competitor could come to Uncle Sam, get access to the same patent, and free-ride off of their investment? As a result, private industry saw academic research as “contaminated” by government funding. With no one willing to invest in commercializing them, the majority of academic innovations ended up gathering dust on a shelf. By 1978 the government had received over 28,000 patents from it’s funding of University research. Fewer than 4% had ever been licensed.
Bayh-Dole brought these inventions back into play with a simple remedy: transferring the rights of patent ownership from the government agency that funded the research to the institution where it was conducted. Universities were now able to negotiate licenses — including exclusive licenses — with whomever they saw fit. And they did. The results were felt across the economy, but most acutely by the biotech industry, where commercialization of academically-discovered drugs skyrocketed.
Over the course of the 20th century, the innovation ecosystem of the US evolved from the bare rock and lichen of the pre-war system, where small numbers of scrappy researchers competed for limited private resources, to a stable, intermediate ecosystem centered around the university and underwritten by the public via government agencies.
- The first shift in the 1940s- from private to public funding of research — stabilized the soil for permanent innovation ecosystems to form around universities, especially in major cities
- The second shift in the 1980s- from public to private ownership of research-derived patents — provided the nutrients needed for the resulting innovations to grow and bear fruit
From these stable intermediate ecosystems a number of niche species, and whole new types of ecosystems, emerged and continue to evolve. Silicon Valley and Cambridge, MA are, of course, the prototypical Climax Communities, where every species of entrepreneur can be spotted performing their courting rituals for venture capitalists (easily identifiable, of course, in their neatly pressed vests). Like rainforests, the availability of resources in these cities — ideas, funding, talent — fuel an abundance of variety and activity that self-reinforces and self-perpetuates. Other ecosystems — each poised to become its own unique Climax Community — are taking shape, if more slowly, in other regions: Chicago, Houston, Atlanta, LA. At the foundations of each of these are research universities which themselves continue to evolve and adapt.
Four Types of Research Questions
In addition to the who, where, and how questions of research funding is what kind of research it should be. The ampersand in the abbreviation R&D suggests two types of related activities, “research” and “development”. In an academic setting, the two parts of this diad are more often referred to as 1) “basic” or “fundamental” research and 2) “applied” or “translational” research. The former conceived of as the search for discoveries, for discovery’s sake. The stereotypical image is of Robert Boyle sleeplessly interrogating the properties of various fluids, Rosalind Franklin patiently compiling x-ray crystallographs, or, more recently, teams of scientists pouring over diffraction patterns of energized particle collisions in search of the Higgs boson. In each of these cases the researcher is driven, at least in our imaginations, by an unalloyed desire to know: why is the world the way that it is? how does it work?
The latter is about applying the fruits of discovery toward solving everyday problems, be those winning wars, treating diseases, improving industrial processes, or just making life easier. Stereotypical images are of Edison sleeping under his desk beside his 999 light-bulb filaments, or of the dozens of scientists gathered in the New Mexico desert to see if their nuclear contraption would really explode. These researchers are driven, again at least in our imaginations, by the desire to solve: how might we create light without oil? how might we end the war without my friends having to die invading Japan?
In practice, however, the distinction between basic and applied research is rarely so neat. Was the Human Genome Project basic or translational research? Was Jennifer Doudna’s lab conducting basic or applied research when they interrogated bacterial immune systems and discovered Cas9? Were their motivations to know or to solve? Does it matter? When thinking about the nature of innovation ecosystems, the distinction between basic and applied research can obscure as much as it illuminates.
A better distinction, perhaps, is between the types of questions that research can address, and the different types of innovations that result. The first two are those most often associated with the modern research university and national labs.
- Knowledge Questions: These involve knowledge gaps, puzzles, or incongruities within a particular field of knowledge. How do different non-coding RNAs modulate DNA transcription? What role does oxygen play in a tumor? Why don’t the observations from our newest telescope align with predictions from the Standard Model? The answers to these types of questions are scientific discoveries.
- Technical Questions: These concern technical challenges that 1) might flow from scientific discoveries or, alternatively, 2) might need to be overcome in order to test theories & make discoveries in the first place. How might we “image” a protein? How might we deliver a nucleic acid into a cellular organelle and detect whether it got there? How might we manufacture, align, and cool superconducting electromagnets? The answers to these types of questions are novel technologies.
Two other sets of questions are often conceived as not pertaining to scientific research at all; but if a goal of the university is not just to generate discoveries but to facilitate their translation into solutions to real human problems, they are just as important to consider and get right.
- Customer/Problem Questions: These start with specific customer problems and work backward to technologies that might solve them. The most obvious place to find these at a university is in an engineering college or hospital-adjacent department. How might we desensitize children to food allergies? How might we switch tumors from cold to hot? How might we decrease the negative side-effects of radiation therapy? How might we reduce the amount of greenhouse gas that heavy manufacturing generates? The answers to these questions are potential new products.
- Commercial Questions: These include the host of practical questions related to founding, capitalizing, and operating a new venture. These are the questions asked in business schools and MBA programs. How do I build and protect a sustainable business model around a therapeutics platform? How do I raise money for a device company? How do I attract an experienced CEO to a seed-stage company? The answers to these questions, when answered together, are new ventures.
To be able to effectively ask and answer questions in any of these categories — scientific, technical, customer/problem, and commercial — requires a combination of tools, talent, and training unique to its type. Scientific questions in the biological sciences, for example, require wet labs, complex instrumentation, cell lines, engineered animal lineages, and teams of highly-skilled scientists and technicians. These, in turn, require substantial investments in both physical and intellectual infrastructure that can take decades to tune and mature. All that is to say that very few geographies, let alone universities, have the resources to be able to tackle all four sets of questions well. Like in the natural world, most ecosystems rely on other ecosystems to supply some important input or take in an output. And that is not necessarily a bad thing. It is only in climax communities that the components needed to answer all four types of questions are, more or less, present.
The Next Phase
The last ten years have been particularly exciting for early-stage innovation. It has become increasingly clear that if research universities are to succeed at creating real world impact from their discoveries, they need to go beyond the knowledge and technical questions they are used to answering and become capable navigators of product/customer and commercial questions as well. Like all important realizations, this one happened “gradually, and then suddenly” with a few institutions slowly leading the way, then going all in, and everyone else following suit. By my reckoning, in just ten years every major research institution has either launched, re-launched, overhauled, or significantly expanded its innovation and entrepreneurship center (IEC).
The genesis, design, location within the university, and mandate of IECs differed widely at first. Some focused on students, others on faculty; Some bolted on to an existing tech transfer office; others were built up alongside it, and others were created to replace it outright. Some spun out of a professional school, while others out of an academic division. But like all new organizational models (or products or services), what began with a period of experimentation and proliferation eventually converged and consolidated to a more-or-less standard model across universities, at least at the top tier. The current model for university-supported innovation tends to encompass four main themes: 1) developing & protecting a technology, 2) securing talent, 3) raising capital, and 4) building a business, with support for each spanning from the initial “Aha” moment in the lab or dorm room to the creation, and sometimes early operation, of a company.
Perhaps more interestingly — and of course I’m biased here — is the emergence of new organizations around the University: new forms of capital, both dilutive and non-dilutive; new organizations to train and match talent; even new types of physical infrastructure. Breakout Labs, CRADL, JLabs, Nucleate, and Portal Innovations typify the depth and diversity of organizations that have evolved over just the last few years to fill important niches in early-stage biotech. There is a lot happening in early stage ecosystems. Knowing how to navigate them — who to talk to, where to find resources, where not to waste your time — is more complicated than ever.
As the name of this series implies, my goal is not to cover innovation from soup to nuts, but from “Aha” to Series A: the earliest stages. Much of this time is spent on campus, so I’ll continue spending much of my time looking at universities and, just beyond them, at investors and incubators that specialize in transitioning ventures from university research labs into labs of their own.
