The Productivity Commission on Patents – It’s All About ‘Value’ and ‘Quality’

The final report of the Australian Productivity Commission (PC) on its Inquiry into Australia’s Intellectual Property Arrangements (‘IP Report’) has now been released.  It is quite a tome, coming in at over 750 pages!  I wrote on previous occasions about the recommendation in Chapter 8 of the earlier draft report that computer software be excluded from patentability, and about appearing before the Commission to make submissions opposing this recommendation.

In view of this, the first point I want to note about the IP Report is that the PC heard what I, and a number of other contributors, had to say on this subject.  The draft Chapter 8 has been completely reworked in the final Chapter 9, and the proposed recommendation has been discarded.  In its place. the Commission proposes the more cautious course of action to allow the full impact of recent court decisions and other changes in the law to be assessed, and additional data gathered, to inform future studies and policy-making with regard to the role of patents in relation to computer-implemented innovations.

Other conclusions and recommendations of the IP Report relating to patents include:

1. the addition of an ‘objects clause’ to the Patents Act 1990 (Chapter 7);
2. a further raising of the inventive step standard, to equal or above the highest level in the world today (which, it is suggested, is that of the European patent system) (Chapter 7);
3. requiring inventiveness to be reflected in ‘technical features’ of an invention and not, for example, in features ‘that seem trivial or aesthetic in nature, and thus are likely to do little to advance human knowledge or create knowledge spillovers’ (Chapter 7);
4. raising and restructuring selected fees to discourage maintenance of ‘low-value’ patents (Chapter 7);
5. abolishing the innovation patent (Chapter 8); and
6. reforming the regime for granting extensions of patent term for pharmaceutical products (Chapter 10).

Chapter 10, dealing specifically with pharmaceuticals, is quite lengthy, and covers a range of issue that are not purely patent-related, such as data protection for biologics (there is no case, the Report concludes, for any extension to existing periods of protection) and ‘pay-for-delay’ arrangements (where it is recommended that the Government introduce a formal system for transparent reporting of settlement agreements between originating and generic manufacturers).

All of these recommendations, it should be noted, will ‘weaken’ the patent system, when viewed from the perspective of patent owners.  That is, they will make patents more difficult and costly to obtain, enforce and maintain.  Conversely, the PC would regard such reforms as ‘strengthening’ the patent system, in terms of its wider social and economic ‘value’.

I expect that some practitioners and users of the patent system (i.e. attorneys, inventors and applicants) will be bemused by many of the IP Report’s conclusions and recommendations, and perhaps attribute them to the rarefied approaches that we have come to expect from economists.  If so, then to some extent, at least, they may be correct.  But that is no reason to avoid trying to understand where the PC is coming from.

And where the PC is coming from is its perspective on ‘value’ and ‘quality’.  But that may not be ‘value’ and ‘quality’ as you know them.  When an economist uses these words, they usually mean something quite different from what a patent owner or patent attorney would probably understand them to mean, as I shall explain.

A Reprieve for Patents on Computer-Implemented Inventions

Having been critical of the draft recommendation to abolish ‘software patents’, and (along with a number of others) made written and oral submissions to the Inquiry on this issue, I would first like to say a few words about the significant shift in the PC’s position between the draft and final reports.

Many of the issues that were raised – such as the inappropriateness of bundling ‘business methods and software’ into a single monolithic category, the impact of the Full Federal Court decisions in Research Affiliates and RPL Central on patenting of computer-implemented business methods, the wide-ranging nature and deployment of computer-implemented technology, the inadequacy of copyright alone to protect innovative functionality, the longevity of at least some software-based inventions, and the importance of patent protection as a signal to prospective investors in new ventures – are reflected in the final report. 

The PC, to its very great credit, even appears to have obtained new data of its own volition, and conducted additional analysis in order to test the claims made by the various contributors.  Given appropriate input and direction, the Commission has been able to independently confirm that, in terms of ‘value’, software patents share a similar distribution to patents in all other fields of technology.  It has also been unable to find any empirical evidence to show that other information it received in submissions is incorrect. 

As a result, the IP Report’s final position on software patents is essentially that IP Australia should maintain a watching brief, in order to gather and publish additional data that will be of assistance in conducting analysis and making more concrete recommendations in future.  This is, in my view, and excellent outcome.  A blanket exclusion of patents on computer-implemented inventions is inappropriate and would have numerous unintended consequences.  In the absence of credible objective evidence to the contrary, software-based technologies should not be treated any differently from other technologies under the Patents Act. 

‘Value’ and ‘Quality’ – The Great Divide

As a former researcher, erstwhile inventor, two-time startup employee and patent attorney, I think I have a pretty good handle on what I mean when I talk about patent ‘value’ and patent ‘quality’.

To me, a valuable patent is one that protects some key aspect of a business’ technology, such as a product or feature that provides a competitive advantage in the marketplace.  Such a patent is valuable in the sense that it adds value to the business, and may even be an economically valuable asset in itself (i.e. something that can be sold or licensed for money or other valuable consideration).

Similarly, a high quality patent is one that is ‘strong’ in the sense of being relatively impervious to attack and invalidation.  Generally, this will mean that the technical contribution that is protected by the patent easily exceeds whatever threshold of inventiveness is established by the relevant patent laws.  High quality patents are therefore generally those that have been well-drafted, and protect high quality, or more significant, inventions.

The ideal patent, from this perspective, is both valuable and high quality.  But it is quite possible to own a valuable patent, i.e. one that is, in practice, preventing competitors from encroaching upon a key market, that is not also high quality – at least for as long as nobody decides to challenge the patent’s validity!  It is also possible to obtain a high quality patent that is not valuable, for example because while the invention is in some sense a ‘breakthrough’ it has little or no commercial value.

However, when economists – including those on the Productivity Commission – refer to ‘value’ or ‘quality’, they often mean something quite different.  As far as I can tell, although there may be some subtle difference, the words ‘value’ and ‘quality’, when used in relation to patents in the IP Report, are largely interchangeable.

To an economist, a valuable patent, or one having high patent quality, is a patent that has a high social value.  More specifically, this value is assessed according to how well the patent right is aligned with the stated policy objectives of the patent system.

The PC position on the policy underlying the IP system – in line with the thinking of mainstream economists – is expressed in four principles that it applies across the whole inquiry.  These are:

1. effectiveness — that the IP system encourages the creation and dissemination of valuable ideas that would not have occurred in the absence of the system
2. efficiency — that ideas are generated by the most efficient, lowest-cost creators, traded freely, and do not unduly impede competition
3. adaptability — that the IP system adapts to changes in technology, markets and economic conditions
4. accountability — that changes to the IP system are transparent, evidence-based and reflect community values.

Applying these principles to patents, in particular, a high quality or high value patent is one which:

1. protects an invention that would not have come about but for the availability of patent rights;
2. represents a significant advance over the state of the art; and
3. results in the disclosure and dissemination of knowledge that might otherwise have remained undiscovered or, perhaps, have been retained by the innovator as a trade secret.

These are, plainly, lofty goals.  It should therefore come as no surprise that research cited in the IP Report shows that the vast majority of patents fall far short, or that the logical conclusion is that there should be fewer, but ‘better’, patents.

Understanding the PC’s Recommendations

Once you understand where the PC is coming from when the Report refers to ‘value’ or ‘quality’, it is easy to see how it arrives at its conclusions and recommendations.

The PC recommends that the Patents Act should include an ‘objects clause’, to set out and clarify the underlying purpose of the legislation.  There is nothing wrong with this, in principle, however it does presuppose that we all agree on that purpose.  The PC’s draft report proposed an objects clause stating the purpose of the Patents Act as ‘enhancing the wellbeing of Australians by providing patent protection to socially valuable innovations that would not have otherwise occurred.’  This generated a number of critical submissions which the PC took on board. Thus, while the final Report maintains that ‘targeting socially valuable and additional innovations is a necessary condition for the patent system to meet the principles of a well-functioning IP system’ (which, to my mind, begs the question), Recommendation 7.1 states that:

The objects clause should describe the purpose of the legislation as enhancing the wellbeing of Australians by promoting technological innovation and the transfer and dissemination of technology. In so doing, the patent system should balance over time the interests of producers, owners and users of technology.

With regard to the standard of inventive step, the IP Report concludes that the Australian requirement for a ‘scintilla’ of inventiveness is inadequate (I disagree, principally because I consider the notion of ‘degrees’ of invention unworkable in practice), that the legislative test for inventive step in section 7 of the Patents Act is unnecessarily complex and restrictive (I agree with this), and that a test of ‘obvious to try’ should be considered suitable ‘in some instances’ (broadly speaking, I would have no objections, so long as it is clear which instances it applies to).  The Report specifically recommends further amending the Patents Act so that the Australian test is essentially equivalent to that in Europe.  I cannot argue with this, having previously written almost exactly the same thing myself!

In making this recommendation, however, the Commission has pulled-up short of its apparent preference for Australia to adopt an even tougher, world-leading, standard for inventive step.  The IP Report concludes that other countries, including the US and Europe, also grant too many ‘low value’ patents, indicating that their thresholds also fall short of the ideal, and states that ‘the inventive step threshold — at a minimum — should meet the highest threshold set by any country with which Australia conducts substantial technology trade’ (emphasis in original).

In its Recommendation 7.3, the PC proposes that ‘IP Australia should reform its patent filing processes to require applicants to identify the technical features of the invention in the set of claims.’  This, the Commission believes, would facilitate higher-quality examination, particularly in view of its proposal that:

The inventive step and novelty tests should better target an innovation’s technical features. This would help ensure the patent system targets technological innovations, especially where innovations are based on a mix of technical and non-technical features, or mainly involve obvious and/or trivial combinations of pre-existing objects.

This idea seems to have been inspired by Europe’s ‘technicality’ requirement (something that does not currently exist in Australian patent law), along with the PC’s belief that inventions it regards as ‘non-technical’ are unworthy of patent protection in light of its assumed policy principles.  A list of examples of such ‘unworthy’ patents granted in Australia is provided in Box 7.1 (p 203), including:

1. a candle wick formed with a three dimensional pattern (AU2012351979);
2. a pizza box with outer and inner wall panels, with perforated removable portions in the outer wall (AU2008255242);
3. a set of plant pots that lock together (AU2003202654); and
4. a pet toy with an internal skeletal member and sound-emitting squeaker device (AU2010202886).

Abolishing the innovation patent is, of course, a no-brainer if you use the PC’s principles as a starting point.   If the current inventive step standard is too low and results in too many ‘low-quality/value’ patents being granted, then the innovative step standard, which is even lower, is economically unconscionable – even if the rights granted have a more limited maximum term of only eight years.  Unsurprisingly, therefore – and despite many submissions in support of retaining a (preferably modified) innovation patent system – the Report states:

From a policy perspective however, it is evident that the design features of the [innovation patent system], and in particular the low threshold for the innovative step, contributes to a multitude of low value patents, which can impose costs on the community.

Conclusion – Are We Really Agreed on the Principles?

Overall, it is clear that the Commission’s recommended program of reforms is intended to result in there being fewer patents, by making it harder to obtain a standard patent, impossible to obtain an innovation patent, and more expensive to maintain a patent once granted.  It is very difficult to argue with the recommendations, if you accept the PC’s underlying assumptions about patent ‘quality’ and ‘value’, along with their general principle that ‘targeting socially valuable and additional innovations is a necessary condition for the patent system to meet the principles of a well-functioning IP system.’  On this basis, we would only ever want to grant patents for the very best, most challenging and most socially valuable inventions.  Clearly, however, no patent system in the world meets this objective.  (‘Method of Exercising a Cat’, anyone?!)

We, as stakeholders in the patent system, must therefore ask ourselves whether we consider the PC’s definitions of ‘quality’ and ‘value’ to be sufficient (or even necessary)?  Even supposing we do, are we satisfied that the analysis methods they (and the researchers whose work they cite) have employed accurately measure these quantities?  And do we agree with the policy principles that underlie the entire report?  Because if the answer to any of these questions is ‘no’ then all of the Commission’s recommendations rest on shaky foundations.

Maybe we still want patents for lesser inventions to be available to small businesses, to assist them in attracting capital to bring products to market, employ additional staff, and conduct further (and perhaps better) R&D.  (I say ‘still’ because that was certainly the policy rationale behind the original introduction of the innovation patent.) 

And does it really matter whether an invention would not have come about but for the availability of patent protection?  It is actually quite likely that most commercially-valuable innovations, in most industries, would indeed come about whether or not we have a patent system.  But that is not the end of the story, because we would still need to determine when, where, by what commercial pathways and at what cost that would have happened.  Are we better or worse off because of the patent system that we have, as compared with the ‘ideal’ system that the PC would have for us?  I see nothing in all of the data and analysis reflected in the IP Report that answers this question unless you assume, a priori, that the answer must conform to the policy principles that the Commission establishes at the outset.

The Government is conducting further consultation on the Report, and is accepting new submissions until 14 February 2017.  There’s something for you to do on your holidays!

Tags: Australia, Innovation policy, Law reform, Patent law, Public policy, Reviews