NAME: Ammon Salter
POSITION: Reader and co-director, Innovation Studies Centre, Imperial College Business School; fellow, Advanced Institute of Management Research
In the US tech-transfer community, a common lament among industry participants is that academic bureaucracy and a narrow focus on patenting and licensing often impedes collaborations.
It turns out the US may not have a monopoly on this perception.
Results of a recent survey of UK businesses suggests that many of them are also finding it more difficult than ever to forge university-industry collaborations despite several recent efforts by the UK government to promote such partnerships.
The study, conducted by the UK's Advanced Institute of Management Research and released last week, surveyed approximately 3,400 individuals from 3,100 companies that participated in projects last year sponsored by the Engineering and Physical Sciences Research Council in the UK.
The study also incorporated results from a previous survey of EPSRC industrial collaborators conducted in 2004 by the Science Policy Research Unit at the University of Sussex. EPSRC is the largest UK research council, providing £740 million ($1.03 billion) in research funding in 2007.
According to the study released last week, a majority of surveyed industrial players said they feel that university officials are increasingly impeding collaborations with long pre-partnership negotiations.
They also said they believe their academic counterparts hold unrealistic expectations about the economic value of their IP.
Specifically, about 55 percent of respondents blamed administration and university regulations for the slowdown; while about 49 percent felt that universities "constantly" overvalued their IP.
The full report, entitled "The Search for Talent and Technology," can be downloaded from the AIM website.
This week, BTW interviewed one of the survey's co-authors, Ammon Salter, an AIM Fellow and researcher at the Imperial College Business School in London, about the results of the survey and the state of industry-academic relations in the UK and US.
Following is an edited transcript.
Your study surveyed public-private partnerships mostly in engineering and the physical sciences. Do you think that the findings hold true for the biomedical and pharmaceutical arenas?
Some of the firms in the sample are chemical or biochemical companies. The survey also included some large pharmaceutical firms such as Pfizer and GlaxoSmithKline; and several other large pharmaceutical firms are involved in research collaborations through the EPSRC.
I think there is relevance between what we found in our report and life sciences in general. The major difference relates to IP, which in the life sciences tends to focus on things like a molecule, so it can be more difficult to initially define the IP than it is in physical sciences.
But we got a sense that these concerns are widespread across industry collaborators and not just confined to engineering and the physical sciences.
[ pagebreak ]
There has been a lot of dialogue in the US about a disconnect between industry and academia in research collaborations, but there is a perception that this hasn't been as problematic in the UK. Does your research indicate otherwise?
Over the last 15 years or so the UK has been influenced by the experiences of the Bayh-Dole Act in the US. The government has encouraged the growth of a tech-transfer industry inside the university sector, to try to replicate some of the experiences that the US may or may not have had.
Part of that effort has been to invest in and build a tech-transfer function in universities. Some universities in the UK didn't have a tech-transfer office, and some did, but usually only had a few people working in them. Now there are 10 or 20 people working in these offices, so the UK has become very active in this space.
In some ways the UK bought a particular notion of what the US model looks like. It may be that the model that works best for life sciences – where there is a relatively defined piece of technology that arises from a project, and it is patented, and either licensed out or becomes the form of some kind of a spinout venture with VC funding – doesn't work universally.
Of course the US experience is complex. There is a lot of research done on university patenting in the US in the last 20 years or so. And over that time, though the number of university patents has gone up tremendously, the quality of those patents has declined in terms of the number of citations they receive from other patents.
In addition, the growth in patenting started before the Bayh-Dole Act. And the patenting is concentrated heavily in certain fields, such as life sciences. We find a similar pattern in Europe: we had a growth of patenting in the life sciences prior to US-influenced models of tech transfer.
One can say that the European experience has been heavily influenced and driven by one version of the US model, and this has had tremendous implications for the way university researchers work here. Prior to recent developments, we had a very informal system in the UK where individual academics could sign confidentiality agreements with research partners. Now that's centralized in the university system. There has been an increasing level of involvement by university administration and tech-transfer offices in the process of exchange. In the past these exchanges took place largely informally over a handshake or a beer at the pub. As things have become more and more formalized, that has caused concern.
I think that what we picked up in the survey is that there is a widespread belief among UK businesses that it has become more difficult for them to work with universities in the last four or five years, partly because there are all these new actors in the system, each of which has something to say about the exchange. Conflicts over IP can lead to long periods of pre-project negotiation. I recently read about an example of a company that spent almost 18 months meeting with researchers and IP departments in order to launch a project with a university research partner. I've spoken with people in the pharmaceutical sector that have had similar experiences.
It's partly that it is so challenging to define IP [ahead of time]. When writing a contract for a piece of research, you don't know what the follow-on benefits are going to be or what you're going to find before you do the research. And bringing another player into the game in the form of a tech-transfer office or university administration really complicates things.
Did you make any distinction between difficulties in industry-university collaborations and industry-university licensing deals?
No, we didn't look at licensing, per se. The interesting thing about this survey is that it is two periods of time. Over the last five years, the number of firms who say that the tech-transfer offices have an overvalued sense of their own university's IP has doubled. It used to be that 24 percent of firms thought this was the case, and now it is 50 percent. The percentage of firms that think university regulations and administration are barriers to interaction has also doubled. This is a rather short time period to see these types of increases. It may be that there have been changes in the universities' policies, but I think the academic system is shifting more toward at least one model of the US style of IP-based interaction. And that's a cumbersome system for many of these firms to get used to. It might not be as problematic for life sciences firms, which are more used to these types of interactions, but in areas like engineering, they're completely new.
[ pagebreak ]
Do you think that legislation like Bayh-Dole helps justify the existence of tech-transfer offices in the US, but that because the UK does not have similar legislation that the offices are not as necessary there?
We have a similar framework to Bayh-Dole here in that universities own IP arising from publicly funded research. The difference is – and I think this is true in the US, as well – if you think about how academics engage with industry, there is a broad menu of activities they are involved in, some of which lead to formal IP, such as in a joint research project, which may lead to a new molecule.
But a lot of the interactions don't involve that type of IP. The problem is, when you start imposing the same requirements as if you're assuming that every interaction involves the creation of some valuable IP, then all the interactions start to get clouded by this notion.
It's not about whether Bayh-Dole is good or bad; it's about what the implications of Bayh-Dole are for the continued pattern of interaction between universities and industry, both in the US where the model was born, and in the UK where the model has been adopted. What's happening in the UK is that we've become convinced that there is all this technology lying on the shelves at universities, and the goal of the tech-transfer office is to unlock that technology.
In fact, there may be other mechanism by which that technology gets used and disseminated, such as the training of skilled PhD students, who then go off and work for industry. There may be more effective mechanisms for tech transfer than simply trying to control and protect everything through formal IP and then license it out.
Have you looked at what other alternative mechanisms might work?
There are other interactions between industry and university that are more important than the formal ones – student education, training, doctoral education, and all the rest are very important elements in the exchange between university and industry.
If you actually look at the income that UK universities receive from IP, it is quite small when compared to other forms of income. For example, the total income from IP in the whole of the UK university sector is £38 million. The amount of money it gets from consulting is £288 million. So there are other mechanisms of exchange such as contract research and consultancy that are very powerful drivers of exchange between university and industry, and are three, four, or even five times as important a factor as IP.
Of course, this is true for the US, as well. If you look at the role of IP in funding the University of California, it's quite small. So the question is whether a focus on IP undermines other forms of interaction which may be very beneficial, but may not be captured through formal interactions such as patenting and licensing. The fear in the UK is that we've gone toward a system that involves more and more protection, which may undermine other forms of informal interaction.
The goal of Bayh-Dole was not to generate huge amounts of income for a university; it is a way of transferring technology from universities to industry. If the university adopts a strong view of IP, they may generate other problems, especially on the industry side in that interactions with universities are now very difficult. There is a process of readjustment as industry gets used to working with university tech-transfer offices, which may be a very different mode of action than they've been used to in the past five or 10 years.
As you pointed out, the idea of university tech-transfer offices is still very young, even in the US. Do you think the industry-academia chasm could be the result of a lack of experience or training within tech-transfer offices?
This could be, but UK university tech-transfer offices worked hard to recruit people from industry.
I think one thing you have to separate out here is that the universities in the UK are driven by incentive systems, by the government, to capture IP, and are measured by the number of patents, licenses, et cetera. And these mechanisms actually determine a level of funding at UK universities. So there are very strong incentives for universities to be active in this area.
And there are other things they may be doing that they don't get incentivized to do, so they may be put aside. In terms of staffing tech-transfer offices, it will help to have a broad base of industrial skills. But I wouldn't say tech-transfer offices are lacking in some of these skills. I think it's a very difficult task for these offices. Their goal is to capture wealth from IP developed by the university. And that goal may interfere with or make it difficult for other forms of interaction.
Your survey indicates that industry respondents ranked Imperial College London as one of the easiest universities to work with.
Well, we asked firms to indicate their five most important collaborators on the university side. We didn't ask which ones were easiest to work with or least bureaucratic, though that might be an interesting question to ask. These are the kinds of questions we need to start asking industry, because they need to start thinking about who they want to work with in terms of ease. Universities haven't traditionally thought of this area as one of their mandates.
And Imperial College was first on that list. But it has a fairly large tech-transfer company that might suggest bureaucracy.
We have an organization called Imperial Innovations. But Imperial College and Cambridge [which was second on the list] are some of the leading research institutions in the UK. Something like 10 percent of total research expenditures in the UK are done at Imperial College. And that makes us a very attractive partner, because we know lots of things, and we have a very bright and able faculty and staff. I don't think the quality of the TTO is not a major concern for industry when collaborating with universities. They're looking for really good researchers that are leading their field and can plug them into a research network that they can't get to themselves.