Start ups in the Silicon Cape?

Written by Andrew Rens on October 15th, 2009

Its commonplace for South Africans who visit San Francisco to be told that it resembles Cape Town. Since I lived in San Francisco first and now live in Cape Town I can confirm that Cape Town resembles San Francisco. At least it resembles San Francisco in some respects; both have locations of incredible natural beauty, both are favoured tourist destinations with storied histories, and both are magnets for a wide variety of artists, techies and unusual and colourful characters. There are certainly just as many differences, Cape Town has embarrassingly poor bandwidth, breathtakingly poor road planning, a paucity of public transport and a surfeit of crime.

Despite these differences “the idea that the Cape in South Africa had the potential to become the Silicon Valley of Africa” has captured the imagination of a group of people who’ve formed the Silicon Cape initiative.

“The Silicon Cape vision is of an ecosystem in the Western Cape of South Africa, that serves to attract and bring together local and foreign investors, the brightest technical talent, and the most promising entrepreneurs, to foster the creation and growth of world-class IP start-up companies in an environment that competes with other similar hubs around the world against the backdrop of one of the most beautiful settings and pleasant places to live, work and play on the globe.”

While that may seem somewhat idealistic I must give the organiser’s credit for not claiming to own the idea, the website states:

“The concept of the Silicon Cape is not owned or controlled by any single entity. It is a living community and an organic, ever-developing concept in the hearts and minds of every participant…”

I’d classify myself as a critical friend of the Silicon Cape idea. However I am concerned that in addition to the differences which I’ve listed, there is one critical difference which could founder the Silicon Cape. The San Francisco Bay area includes quite a number of very good universities, not least the public school, Berkeley and the private school Stanford. The creation of Silicon Valley is credited to the proximity of these schools, and the interplay between the faculty and postgraduate students and the entrepreneurs who turn the knowledge from those institutions into wealth producing businesses.

The kind of interplay which I witnessed in San Francisco will be stifled by the provisions of the Intellectual Property Rights from Publicly Financed Research Act. The Act, which is not yet in force, has consequences for South African entrepreneurs interacting with universities and research councils. [Edited to add: I have tried to keep my explanation simple and jargon free. In response I have received criticism from a devotee of the scheme created by the Act. I have inserted a few amendments to clarify points in italics and set out longer technical explanations in footnotes, so as not to break the flow of the explanation.]

The Act automatically, by operation of law(1) gives 100% ownership of any intellectual property which was developed with public research funds to the government (2). The Act does not set a threshold for how much public research assistance someone must get to fall under the Act, as a result if an entrepreneur gets 0,01% of her finance from government research funding then all the intellectual property belongs to the government. This assistance need not be money but could be office space or use of a laboratory or even travel money to attend a conference (if it comes from a research fund).

The Act does allow a university to enter into an agreement changing the automatic outcome of the Act which gives the university all the rights, to give co-ownership to an entrepreneur. Co-ownership doesn’t however give an entrepreneur much freedom of action, since most acts taken in respect of intellectual property require the consent of the other co-owners given them veto power of an entrepreneurs business plans.(3)

Entrepreneurs who are concerned about their intellectual property falling under the Act should be extremely careful when working with universities and research councils like Mintek or the CSIR because the proponents of the Act claim that it applies automatically to any person receiving funding. The most important thing to ascertain is whether there is research funding involved and whether the entrepreneur can be regarded as having received any assistance from it.

Proponents of the Act will tell you that the Act requires that universities (and the new agency created to hoard patents) must give licences to SME’s, and the Act does say that, but the SME’s can’t in the first instance get exclusive licences, so incumbent companies will have access to the same technology. In other words entrepreneurs will be able to use technology based on government funded research but that technology must not be an entrepreneurs unique value proposition because it will be available to everyone else. While an incumbent corporation won’t be able to do an exclusive deal with a research institution and lock up technology (which is good for start ups) an entrepreneur will not be able to rely on intellectual property from a research institution as its unique value proposition. Its not too great a concern to an entrepreneur whose unique value proposition is speed to market or low price but otherwise it should be a large red warning light.
The Act does permit institutions to exceptionally give exclusive licences but the granting of such a licence is in the gift of the university, it must also be justified to a government department and the exclusive licensee must manufacture or offer the services in South Africa. Can you imagine such exclusive licences being granted quickly?

More profoundly choosing to work with a university or research council means that the architect of the Act has chosen an entrepreneur’s business model for her. The Act makes it difficult at best to develop FLOSS, crowd sourced or other open licensed business if one has any public research funding. Whether it will be possible at all depends on regulations which have not yet been finalised. The Act also prevents an entrepreneur from owning the intellectual property her business is based on, and makes obtaining an exclusive licence difficult. In other words it eliminates both ends of the spectrum, open businesses and businesses based on a monopoly on a particular technology.

Another more diffuse effect of the Act on entrepreneurs will be to reduce the knowledge available to entrepreneurs. Journal articles have been becoming increasingly available online in open access journals as the result of a policy by the same Department (Science and Technology) which sponsored the Act. The Act though prohibits publication of new knowledge until patent priority is obtained. Unlike open access journals the full text of South African patents are not available online, only a type of summary is available online and then only if one pays a subscription fee to the database. The database lacks state of the art search functionality. As a result entrepreneurs in technological fields will not be able to efficiently make use of taxpayer funded research.(4)

A researcher is allowed to publish an article after the patent is granted but the researcher would have had to wait while the university decided to patent or not, and then if not continue waiting for a government office to make the same decision. If a patent is to be applied for then the researcher must wait for the application, by which time someone somewhere else in the world is likely to have published first and the research is no longer cutting edge.

The Act is not yet in force and it could be amended.

(1) The way that the Act is worded it assumes that where there is any contribution by public research funds to intellectual property that all of that intellectual property is to be owned by the university or the government. If an entrepreneur pays 99,99% of the research costs and public research funds pays 0,01% then 100% vests in the university or the government.This result arises ex legefrom the operation of the law. Since it appropriates the money and effort of others and deems the intellectual property to be 100% publicly financed even when it is not, it is probably unconstitutional. It can be altered by a contract entered into beforehand by an entrepreneur and a university, but only to give co-ownership. Entrepreneurs need to be alert because if they don’t enter into a specific agreement that alters the automatic operation of the law they will lose all intellectual property rights.

(2) A commentator has objected that universities are not the same as the government but autonomous. We can certainly hope that this is true more generally since a number of universities fought hard during apartheid to be autonomous, and academic freedom is included in the Bill of Rights. Do universities have any autonomy under the Act? They are required to created technology transfer offices who are legally responsible directly to a government department, bypassing university leadership, those offices are to be micromanaged by the National Intellectual Property Office which can second guess their decisions, and government can take over the intellectual property in terms of a number of different provisions of the Act and regulations.

(3) The Supreme Court of Appeal recently decided a case on co-ownership of copyrihgt Feldman NO v EMI Music SA (Pty) Ltd/ EMI Music Publishing SA (Pty) Ltd (268/2008) [2009] ZASCA 75 (1 June 2009). The court held that a co-owner cannot sue for copyright infringement without the consent of the co-owners. The court also suggested, although it was not directly in issue before the court that “In the absence of clear contrary indications as to the parties’ intention, it is suggested, the co-authors will each be taken to hold an equal, undivided share as ‘tenants in common’. If that is correct, and the Supreme Court of Appeal is the highest court on non-constitutional issues, then a co-owner needs the consent of the other co-owners to licence the work, to set royalties, to collect royalties etc. In those circumstance an entrepreneur cannot act independently or quickly.

(4) It was suggested that the problems with the South African patent system are not the fault of the Act, however the Act is not a person, its a legislative instrument operated by the same State expending a great deal of resources on setting up the various bodies required by the Act. One of the reasons given for not reforming South Africa’s patent system is a lack of resources, yet resources are being used to create an whole new body to administer intellectual property even though one of the key institutions on which that new body will rely is well known to be inadequate. It would be better to first fix the basics, i.e. patent administration, before embarking on complex schemes like the Act.


2 Comments so far ↓

  1. Jaci Barnett says:

    I am astonished that the author of this was in the same workshop on the regulations as I was a week ago. It appears that he was on a different planet, not just in a different workshop.

    These types of ad hominem comments aren’t appropriate on this blog, and won’t be approved in future.

    I point out some of the more obvious inaccuracies:

    It seems that the reason for the different perceptions is that the blogpost is based on a close reading of the law i.e. what the Act and draft regulations actually provide and not on projections of how the Act might work in some circumstances.

    “The Act gives 100% ownership of any intellectual property which was developed with public research funds to the government.”

    Two issues:
    1. Universities are NOT the government. I am not a lawyer but I understand that although we may be publicly funded and are established by an Act of Parliament, we are autonomous.

    What autonomy? The blogpost now has a footnote discussing how the Act deprives universities of autonomy, probably in an unconstitutional way.

    2. 100% ownership is NOT given in the case of co-funding – then the IP is jointly owned.

    Read the Act which automatically grants technology transfer departments 100% ownership of intellectual property if there is any public research financial contribution. The IP can be jointly owned only if there is a prior agreement which says otherwise. Entrepreneurs are entitled to be alerted to the automatic operations of the Act which may deprive them of intellectual property.

    His statement will lead to the assumption that any IP developed with public funds, even if co-funded only, belongs 100% to the university. This is not true.

    That is the assumption which the Act makes, that any IP developed with public [research] funds belongs 100% to the university or the government.

    However, in the case where only public funds are used, why shouldn’t the university (not the government) own it 100%?

    But that is not how the Act works, instead where public research funds (not just any public funds) are used, even if it is only 1% then the technology transfer office or NIPMO becomes the 100% owner ex lege. Is that equitable? Is that useful? TTO or NIPMO is deemed to be the 100% funder even when it is not.

    Who should own it?

    The entrepreneurs who contributed to it, and who are after all entrepreneurs so its their job to innovate. Innovate is used here in the sense used by Peter Drucker, to actually change how markets operate by introducing new products and services at scale, and not in the wholly mistaken use of the term to mean ‘to register a patent’.

    “…SME’s can’t in the first instance get exclusive licences, so incumbent companies will have access to the same technology…”

    This is NOT what the ACT says. Non-exclusive licensing is preferred by the Act, but exclusive licenses are possible and, certainly, will probably be the way forward for most deals.

    Hence the term first instance.

    “The Act also prevents an entrepreneur from owning the intellectual property her business is based on, and makes obtaining an exclusive licence difficult.”

    This contradicts the earlier statement about the fact that SMEs can’t get exclusive licenses but, if we ignore that contradiction, the Act does not PREVENT an entrepreneur from owning the IP, it is just not a preferred route.

    The Act does prohibit an entrepreneur from 100% owning the jointly developed intellectual property and thus having the freedom to innovate, to move at the speed that markets require. There is a narrow exception that if an entrepreneur can claim to have funded the research 100% (not defined), and both the university and NIPMO both agree then she may obtain exclusive ownership.

    “The Act though prohibits publication of new knowledge until a patent is obtained.”

    This is very misleading. Two issues:
    1. The Act does not prohibit the publication of new knowledge, just KNOWLEDGE THAT CAN BE PROTECTED AND/OR COMMERCIALISED – not all new knowledge is commercialisable and the new regulations are quite clear on what should be done in that case.

    The Act defines intellectual property as subject to intellectual property protection anywhere in the world. In other words if something can be “protected” anywhere in the world then it must be. That covers a lot of categories not protectable in South Africa.
    The Act defines commercialisation as “the process by which any intellectual property emanating from publicly financed research and development is or may be adapted or used for any purpose that may provide any benefit to society or commercial use on reasonable terms”.

    Show me the academic who admits that her research is completely incapable of being used for any purpose that may benefit South African society and I will show you the academic who is unlikely to receive any more public research funds.

    2. The Act prohibits publication until a provisional patent APPLICATION has been made. Does the author even know the difference between OBTAINING a patent and making a patent application.
    This apparent problem was a mere typographical error now corrected
    If we had to wait to publish until we had obtained patents, no publications would ever be made as the process can take many years (there is a patent backlog at the USPTO of 5 years…)
    Its useful that you mention the patent backlog, since patenting (together with registering designs or plant breeders rights) is the only way which the Act envisages to transfer technology. The Act treats global practises in which technology is transferred through teaching,the public domain, joint research, open standards, open licences and participation of academics in entrepreneurial activities as special cases requiring exceptions and in some cases special permission from a government office.

    “Unlike open access journals the full text of South African patents are not available online, only a type of summary is available online and then only if one pays a subscription fee to the database. The database lacks state of the art search functionality.”

    This is hardly the fault of the Act.
    This is issue is now dealt with in a footnote. It does however render the Act ineffective
    This has been raised as a concern for some time and, I understand, is something that is being worked on.
    South Africa’s intellectual property system requires a major overhaul, and until that has been effected legislation such as the Act is at best premature. Vague statements that something is being worked on cannot cure the problem. One might wander why the urgency to put the Act in force and no comparable urgency to fix the patent system.

    “A researcher is allowed to publish an article after the patent is granted but the researcher would have had to wait while the university decided to patent or not…”

    My point above refers. A researcher is allowed to publish after a provisional patent application has been made – NOT GRANTED. I make these decisions and organise a provisional patent application when necessary within 2 days. Hardly a delay for publication.

    What one office is able to achieve before the Act comes into force isn’t a useful indicator of the situation when the Act is in force and all the technology transfer offices are required to patent almost all research thus creating bottlenecks at both technology transfer offices AND at the patent office

    I understand that the author is pushing a particular agenda and he is entitled to his opinion.
    The agenda of this blog has throughout the drafting of the Act and regulations been to analyse the adverse consequences for research, and thus for South Africa’s development. This post though has not focused on those impacts but is rather in response to a number of queries about the impact of the Act and regulations on entrepreneurs who want to know how this will affect their interaction with universities. Thus although this blog usually advocates use of open licences that is not the only business model considered in the post, instead other business models are considered such as those in which exclusive ownership or control of intellectual property form the unique value proposition of an entrepreneur. Its not the agenda of this blog to promote such business models, but if the implications of the Act are to be considered for entrepreneurs then that is an important consideration.
    However, I would expect him to at least get his facts right when he provides his opinion, otherwise we should look at his entire opinion as suspect.
    The blogpost is based on what the Act and the draft regulations actually say, not how the scheme will be marketed to entrepreneurs.

  2. Justin Stanford says:

    A very interesting and very much underexposed issue facing the Silicon Cape ideal, and one that has been raised by a few people in the know. Certainly not a good thing.