Comments for A shorter patent term for some innovations?

The pace at which new software versions come on the market – each three years on average – would be a reasonable time for copyright to last.

I can see several reasons for it:
– Companies need to keep up-to-date their products with the trends of the market and fight obsolescence. They want to attract new consumers and convince old ones that the new version is safer and better. These factors allow them to stay competitive towards their competitors.
– New uses and habits flourish along customers. These customs have to be incorporated in the new versions.
– Companies also have to stay at the competitors’ level, in order not to be left off.
– In the software market, it becomes easier to copy an innovation without falling into the reach of its copyright.

Furthermore, nowadays, in the computer and software sectors, a period of 3 years is a (very) long time. In several months a computer or a program could be outdated.

All these reasons drive innovative companies not only to invest in R&D but also to file copyrights in order to keep a short advantage towards their competitors.

So, even if the copyright length is longer than 3 years, I would recommend the authorities to reduce it.

I would like to talk about the global patent system and the process of how a patent is legally reviewed. According to Stanford University analysis, companies have spent around $20 billion on patent litigation and purchases in the last two years and Google and Apple have spent more money on patent law suits and purchases than on R&D for new products.

In the recent patent war between Apple and Samsung, the two companies have been suing each other over patent infringements, allegedly committed by the opposing party in the context of recently developed smartphones and tablets. The patents held by apple and apparently infringed on by Samsung include both design and utility patents and those held by Samsung are mainly utility patents. The conflict between the two companies has been carried to around 10 different countries through out the globe with varying rulings in different countries. In the US Apple managed to achieve a verdict largely in favour of their interests, whereas in South Korea and Japan the result was more in favour of Samsung. Now, is it surprising that each company won major cases in their respect country of origin?

First of all I would like to take a small look at the patent system on an international basis. Since the establishment of the Patent Cooperation Treaty (PCT) in 1970 and somewhat more unified system has been introduced globally to facilitate the application of patents in different countries. The applicant submits its patent, which is then analysed by an official body. The initial application is then followed with a regional phase, during which the applicant has to apply for a patent in the specific countries. In effect there is no such thing as an international patents, but the system grants various benefit, such as a priority status for the regional phase, after successfully completing the international phase. Having different patents in different countries, then obviously results in different law suits in different countries. The initial aim of the suing company is to achieve an injunction on the sale of the infringing product, after which the suing party seeks to receive a money payment for resulted damages.

So the aim is to avoid conflicting verdicts of different international judicial bodies on the grounds of national interest and to significantly decrease litigation cost for innovative companies.
I can think of two possible solutions for such a system (1) a further development of the PCT into a system of global patents or (2) a unified and global system for reviewing legal claims on patents or an “international court of patents”. In order for (1) to work differences in the process of granting a patent, such as between the US and the EU regarding software patents, must be reviewed and worked out in favour of a multilateral solution. (2) would require the establishment of an international institution and the relegation of a certain degree of sovereignty by the participating countries regarding patents, which might pose a problem.

Patents exist since the 15th century – they were created to incentivize innovation by assuring returns sufficient to recover R&D costs and grant profitability to the inventor. Their strength is that anybody who is able to translate an idea into practical and useful terms can be rewarded with a patent (if he/she is quick enough to ask for the patent before having someone stealing his/her invention). And this happen in all sectors. Which is great, because it guarantees that innovation takes place everywhere.

The issue of IT patents became extremely relevant when the mass of consumers started using computers in their offices and then in their houses. Demand for hardware and software boomed, and producers had to keep innovating to satisfy consumers’ thirst of new gadgets and applications.
Since a patent-free world would lead to insufficient innovation, I believe the European Patent Office does a good job in protecting innovation. In fact, it grants a patent only to real novelties, to those inventions that are not just a “copy and paste plus some improvements” of old ones. By securing protection for 5 years, EPO would make sure new technology is produces, R&D costs are recovered and consumers could soon maximize their surplus. It is true that software needs months or years to reach the mass, but usually it does when it is under copyright/patent protection. Thus, consumers would have their first experiences with that original product and they would set their standards on it. Consequently, when the patent expires, those who want the original program will still purchase it at a high price (because of the availability of continuous updates which increase the value of the software), while those who are satisfied with substitutes will get them almost for free. An example is provided by Adobe’s Photoshop – no matter how many similar products exist in the market –such as Picasa, Canvas and Picnik – people will still buy Photoshop, because it has become the reference program. The fact that in new dictionaries it is possible to find the verb “to photoshop” is an additional proof of it.

To conclude: the IT sector has been expanding for years, and it is not going to stop soon. Thus, it is right to protect innovations so as to foster new ones, but patents should not last for long (5 years might be a good amount of time) to keep up with the quick and massive changes that characterize the sector.

A patent is a form of intellectual property. It consists of a set of exclusive rights granted to an inventor. In my opinion, a patent is often necessary if we want to promote innovation. It is essential to motivate the development of technological inventions and the publication of innovative ideas. But to invest in an innovative project, the inventor must have a positive expected profit. This justifies the presence of patent to safeguard the initial investment. However, the patent has its positive and negative sides.
On one hand, create new things often require a long period in which the inventor receives no income as a result. So, a patent is necessary if you want to reward people for their innovation because they work usually intensively on a project before seeing any benefit effect.
On the other hand, too much protection reduces competition because it prevents the entry on the market by other rival firms. Furthermore, it’s not efficient if the innovation stays trapped with only one inventor. Everybody should be able to take advantage of the invention, even if the creator is not in favor.
The dilemma at the creation of patent is to balance those two competing goals thanks to a time optimal protection

As regards the software patents duration, the optimal period of time is a rather subjective criterion. A too short timing can be harmful to the inventor because the return on investment may be insufficient and therefore prevents innovation. To reach efficiency, firm should price at the marginal cost and thus makes zero profit. However, this is not sustainable want firms must have a profit incentive to take the risk of investing in R&D. Besides, if we don’t grant patents to inventors, most of the firms will wait that other firms make discoveries. This problem is very marked in a competitive sector such as computer sector where each discovery helps the other researchers to innovate.
Instead, a too long patent reduces the growth of the economy because ideas protected by a patent cannot be developed further by other researchers. And it is not sensible to offer a long patent (for example 15 years) in such an evolving sector where discovery is rapidly obsolete. In addition, the creators see no incentive to innovate as they enjoy patent protection. This is inefficient for the market. To grow the market needs to be boosted in its mature phase, particularly through innovation. The consumer may also feel a disadvantage because the diversity of supply is restricted due to low investment in R & D. In short, I think a short patent’s duration (two years for example) is enough to boost the innovation because the computer technology is quickly obsolete.

Based on the model of William Nordhaus, see how to fix the optimal duration of patent. The objective is to find the ideal degree of protection to reach the efficient equilibrium. This balance would both encourage innovation without the innovating firms remain on their laurels after their discovery launched. A positive marginal profit is required to egg on innovations. The producer needs a large positive surplus to start highly risked project like software innovation. The expected benefit should increase thanks to the patent because it is a form of entry barrier which reduces competition and allows the firm to retain or to increase its market share. But we can take into account the marginal loss in terms of consumer surplus. The duration of the patent tends to maximize the social welfare. First, a limited duration of patents fights against monopoly power because when patent expires the competition emerges. Secondly, a short-term patent stimulates firms to diversify in order to benefit from a new patent to offset the expiration of the former patent.

In my view, the question of the duration of the patent is not the only issue. The heart of the problem is rather in the patent criteria which must be strengthened in order to avoid too broad patents. Often the patent is so vague that the holder possesses a means of claiming the ownership of another idea that was not originally intended to be patented.
In addition, patent law should be the same all around the world. Laws should be harmonized so that all countries are on equal footing when innovation emerges. In this case, investors would have no interest to invest in innovation abroad rather than in its country of origin to increase their profit. This may be an interesting reflection to understand massive investments in some countries more or less legalized…

Welcome to the new world, now pretend you’re not here!

The Ideals

The intellectual property rights (IPR) discussion is a complex one, a discussion that is not aided by the diverse number of conceptualizations conceived by opposite poles. On the one side you have economists and other theorists arguing that we need to protect the rights of innovators so that we have an incentive scheme put in place for guaranteeing a continuous innovative society, on the other side we have, again, economists and other theorists as well arguing that the IPR’s are killing innovation because no one can try to develop an innovation without stepping on other past patents.

Seemingly, the debate has polarized these different views though I believe we should ask ourselves, do these views actually differ that much?

It is not about single cases such as Disney, Napster, HIV medicine or Hollywood; it is above and beyond that. It seems that both views have societies ability to innovate as their main concern and objective. So if these ideals are in fact perfectly aligned then why is it so controversial?

Which world do we live in?

The controversy derives from different cognitive realizations of how our society is and how it should be.

I would argue that ‘Pro IPR’s’ tend to view the world in a more static context. In their eyes we need patents to uphold the structure of our economy and growth. This results in assigning false scarcity to intangible assets that in reality are not scarce at all. In other words, we give products, that are beyond our economic system, characteristics to make them fit our economic system.

‘Anti IPR’s’ however view the world in a more (for lack of a better word) idealistic manner. They recognize that innovations come from a common platform. This ideal, I believe, is less competitive and more focused on actual growth in a bigger picture. A good example of this is the extraordinary development the operating system Linux has gone through while many in the IT industry consider Microsoft’s Windows to be at a halt. This is by all means open-source versus IPR.

However, the challenges resulting from an open source ideal are incomprehensible. Because, even though we have reached a point in human history that we can mobilize intangible assets we still live in a structure that will not allow for the intangible assets to become a reality without ownership.

The Outfall
So, do we accept that we live in a world that will prioritize ownership and compartmentalized growth because there simply isn’t room for unincentivized innovation? Or, should we try to figure out how a new structure looks where there is room for open source ideals?

This discussion borders on fundamental economic ideologies and is not answerable in a blog comment. I accept that this comment is rather unorthodox and subjective, yet the only point I want to make is that we need to dig deeper in this discussion. It is not about number of years of patents, it is about how we as an innovative society want to develop, how we want to look in 2,5,50 or even 200 years.

In this comment, I would like to express my feelings about the patents politics and the length for software patents I think would be the most profitable for everyone.

At first, as said in some of the previous comments, I think it’s important to have the same rules regarding software patents (and other patents in general). The differences between the US policies and the European ones can lead to some inequalities in the market. For example, what if a US company is able to develop his software because of the patent given while the European one with a very similar invention is not able to because the definition of a patent is different? It can also lead to an exodus from some of companies to the other continent in order to be protected by those patents which weren’t allowed in their home country. Or it will discourage the companies who might have some very good ideas but don’t beneficiate from the patent. Personally, it think the US patents rules are slightly beter because they are more permisive concerning patents, and giving companies patents is a good way to encourage them to create and innovate.

So, let’s just determine how important are patents. On a company point of view, patents are really important because it protects them from being copied and thus it gives them the incentives to innovate without the fear of being plagiarized. Another positive aspect is that it stimulates the concurrence. Everybody have to find out their own way of developing their succes and not juste wait for anyone to find an idea and them take it. However, a patent can also be a brake to equal competition if there are too many patents (see IBM who was the leader in the 90’s while having more patents than any other companies).

But the danger would be to give too many patents and to give them a long duration. Indeed, the balance between use and incentive has to be rightly equilibrated. As said in the Nordhaus model, giving long duration patents would increase the delay of innovations but it will also cause longer R&D costs. Also the consummer surplus will be low as long as the patent goes as they really benefit from the innovations when the patents are over.

Personally, I think something like 3 years patents could be good. It protects the innovators during a reasonnable period so that they can develop their software quietly but it also push them to not take it too easy. After that period, the other companies could use the ideas without being restricted by the patents to develop something by their own way, espacially in the industry of software where progress is continuous. Furthermore, it allows consumers to benefit from these innovations while they are still modern.

Concerning the debate about software patents, copyrights and trademarks I would like to point out, that the Bernoulli Numbers, which are the basis for all computer programs and software, have never been patented.

On the one hand it is irrefutable, that there are difficulties in dealing with patents for software and especially distinguishing between an idea and a concept. Each bit of code could be argued to be potentially patentable and at the same time the dynamic and innovative industry leads to a shortage of available experts, who could evaluate both the concept and value of the code.

On the other hand companies and individuals want to have an exclusive right to sell and trade their software. They put forward the necessity to create investments and to compensate exogenous sunk costs for development.

It occurs to me that the bigger the companies involved in software industry become, the higher is the amount of invested money and the greater are the incentives to gain patents. Nowadays the disputes about patents are raging and even the biggest companies are involved (e.g. Samsung, Apple and Microsoft).

The previous comments have analysed the negative impacts on patents very precisely, concerning market inefficiency of a monopoly market structure, a lack of investments and a dead weight loss for the population. But I would like to stress the fact that patent and copyright is not just protecting individuals and companies but also national economies and trade balances. That is why all MEDC’s are dealing with the political issue of patents and the protection of intellectual property.

In order to answer the question concerning the duration of patents I wondered how a world without software patents would look like. In fact, for many commercial software programs there is a substitute of free software available. Therefore it seems rather unlikely that patents are creating innovation. But it cannot be proved if commercial software was transferred into free software or vice versa.

Based on the software industry today, dealing with progressive and dynamic markets, most innovations are created in teams at modest costs. As a result the cost differences between inventing and copying are not as high as they are for the pharmaceutical industry for example. So there is no need to protect software for a long period of time. I would approve the duration of a patent for 2 or 3 years, keeping in mind the great importance of lock-in effects for business applications in big companies. The investments for retraining and lost working hours are the cost drivers of a software change, which is therefore avoided.

In my opinion the implementation of web 2.0 and the following hype of cloud computing, creates a platform for experts where they can communicate and exchange ideas. The public and often private cloud solutions for software as a service offer a developer toolkit and interfaces to other programs, so that the software can be personified.

With this comment I would like to address some of the issues around the software patent debate, and hopefully give some new insights to the discussion.

Firstly, looking towards U.S. patent laws, the U.S. Constitution states in its patent laws, that patents should “promote the progress of science and useful arts”. By using the framework of the Nordhaus-model the objective would be to determine the right balance between the incentive to innovate and the social benefits from use by consumers and other firms after the patent expires. With patents being issued right, left and center with durations up to 20 years, this allows companies to exploit monopoly power while also hindering competing firms from researching and developing software in the same area. And if you consider how this affects social welfare by exploiting the consumers with high prices, there really is a simple answer. And if you have a look at the recent competition in the smartphone segment with lawsuits being thrown in every direction between competing firms like Samsung, Apple, Motorola and others, you could ask yourself: Does this promote the progress of science and useful arts? Hardly so.

Another issue of how companies can obtain patents for simple software functions is beyond me. The breadths of some of the patents that get approved are simply astonishing, and limits research and development, and innovations in the same area greatly. If you look at Apple for instance, every function they try to incorporate in their smartphones, will be patented, or at least tried patented, to ensure other companies from copying them. But the techniques used in Apple smartphones are not innovations that haven’t been used before. Multi-touch is a great example, and was when introduced for the iPhones in 2007, claimed by Apple as a new invention, even though multi-touch had been in use for decades already by IBM and Mitsubishi. So, with Apple’s thinking, it is okay to steal from others, but not let others steal from you?

Kirby Ferguson, the inventor of the web-series “Everything is a Remix” (1), interestingly addresses the above question in his speech about patents and copyrights at a Ted-convention earlier this year called “Embrace the Remix” (2). He claims that all creative work is to a certain extent based on the work of others. And here it is vital to differentiate between basing your work on others, and simply copying it.

The incentive to innovate will in my opinion be hindered if you issue patents for everything, at such durations as you do today, as you thereby restrain the creative process of innovation. An option would therefore be to limit patents to a duration of only a few (1-3 years), but I feel that traditional copyrights will serve the software industry as good as patenting, as it allows for a more efficient market economically, and promotes the creative process in a greater way.

You can argue that creativity comes from without, and not from within, as we look at others, and we invent and create new ideas based on the work of others. Just like this blogpost is based on the work of others, like Kirby Ferguson (1) and Richard Posner (3), and then remixed with my added thoughts and ideas.

—-
Sources:

1. Kirby Ferguson (2012) Everything is a Remix, Retrieved October 16, 2012 from http://www.everythingisaremix.info/watch-the-series/

2. Kirby Ferguson (2012) Embrace the remix, Retrieved October 16, 2012 from http://www.youtube.com/watch?v=L1s_PybOuY0

3. Richard Posner (2012) Why there are too many patents in America, Retreieved October 16, 2012 from http://www.theatlantic.com/business/archive/2012/07/why-there-are-too-many-patents-in-america/259725/

In what follows, I argue for a very short (although without formal models and actual figures it is difficult to be specific with respect to the actual number of years) duration of software patents and follow the line of reasoning provided by Bessen and Maskin in “Sequential innovation, patents and imitation” (2009). According to their argument, some industries (with the software one being the exponent) are based on sequential innovation (each future innovation builds upon the ones that preceded it, for instance reusing the same code lines) and complementary innovation (each potential innovator has a different research approach and therefore improves the chances of the innovation to be successful).
One important aspect in this context is the entry cost. Even though the innovation cycle seems to be much shorter in the software industry than in any other, the software development still comports significant investment costs, if not in the “equipment” used, certainly in the skills employed. Thus, entry in the industry will not occur spontaneously and will not compete away the entire initial innovator’s profit. Consequently, the first innovators might still benefit from a first-mover advantage, with the returns to innovation being less driven by patents, but by lead time and learning-curve advantages, as outlined also by Levin et al, in their paper “Appropriating the Returns from Industrial Research and Development” (1987).
There exists also a timing problem in the software patents. Usually, software innovations are made available on the market before their creator is granted a patent protection. According to Perchaud (“Software patents and innovation”, 2003), it takes around 18-24 months in Europe to get a patent, by which time the innovation is already outdated or a similar software might have already been put on the market.
Moreover, a distinct feature of the software industry is the “network effect”, which will create a pseudo-monopoly effect (the developed software will self-select according to the ones that manage to acquire the biggest customer base). Even more, once a customer base is gained, companies have at their disposal tools for keeping it – for instance, switching cost or the selling of regular subscriptions instead of a one-time purchase. This is further increased by the interoperability problem – once a client decides to switch to another software, he might need to make further adjustments, also costly, which might persuade him not to do it after all. Throughout the literature, there exist several empirical studies that have shown how companies that were given the strongest patent protection reduced their R&D costs (contrary to what the standard arguments suggest), such as Bessen & Hunt, “An empirical look at software patents”, 2004 (this comes in agreement to what my colleague Roman has emphasized theoretically, that a monopoly structure with no entry threat has less incentive to innovate than a monopoly industry threatened by entry, which we also proved in the first problem set).
Looking at the problem from the other perspective of the market, that of the consumer, the longer the patent protection, the lower the value accrued to customers and, more broadly, to the society, especially in a fast-moving industry like software, where products become obsolete at a high speed.
All of the arguments outlined above can be fitted into the proposed model of “software useright”, which proposes to distinguish between two markets: a market for the rights to copy software (for consumers) and a second market for the rights to imitate software functionality (for further innovations) and for which copyright seems more appropriate.

In my point of view no patent protection is more effective than if there is any kind of protection. What is the main point of a patent? It is the exclusive right of a winning company to be the only one producer of certain goods (in our case software) – so a winning firm becomes a monopoly on the market of this certain software for the time declared by the patent. I would like to point out some problems which are caused by this situation:

1. Market inefficiency – a monopoly market structure with no possibility to entry (due to patent right) is less efficient than other market structures (e.g. monopoly market structure with open entry to industry). The output on this market would be lover than it’s optimal and price would be higher.

2. Lack of motivation to innovate – there would be no or only a little motivation for innovations and improvement of protected software (especially in the case of long-term patents). It would lead to conservation of the market.

3. Huge sunk costs – costs of all opponents who tried to develop the same software are sunk costs after introduction of the patent. Sunk costs are very high in software industry because share of tangible goods are relatively less and intangible goods are almost worthless after introduction of patent.

On the market without patent protection we can expect a typical “creative destruction” – replacing old software by new ones which can lead to better software sold for lower prices.

In my opinion argumentation that firms would have no motivation to invent are not right, because the winning company will be a short term monopoly which has great power to lock-in their customers. Lock-in effect is so underestimated in case of software patent policy although the lock-in effect is very strong regarding software (good example is an accounting and management software – e.g. SAP).

After considering the lock-in effect we can see that short-term monopolistic position can be sufficient for building a strong market position, so there is still big motivation for firms to invent. Winning firm, moreover, would be forced to improve their software product because it would be on the market with free entry. Consequences are that other competitors can still continue with development of their software and try to overcome the winning firm.

In the following I will give my opinion regarding patents and patent durations for software.

Pros: I think, without any patents no one would have the incentive to innovate, because you have to invest a lot of money to find out new systems and technologies. The longer the patent duration is (but only up to a certain level!), the higher is the incentive to invest more in R&D (see Nordhaus model). If there are no patents, other enterprises could use your innovations without having spent any money. The incentive for an enterprise to innovate is to have or to use something, that nobody else has. If you spent a lot of money in R&D, you want to achieve an advantage out of it. So patents function as a protection.

As a result, I think it is important to award patents for software, but it should be discussed how long the duration of such a patent should be. It is a big problem, when patents are valid for 20 years. The computer software industry is a very fast one. It is kind of impossible to predict the next 20 years. Even a 2-year horizon, which should be the longest duration in my opinion, is a challenge.

Contras: To get a software patent is very expensive and thus risky. Most companies can’t afford this risk. The result is that only a few companies (or monopoly) are able to be active in the software domain. As a result, software users have only a very little choice between different products, which give the patent-owning companies kind of might.

Another disadvantage is, that a patented software feature cannot be implemented from other software developers for 20 years (if this is the duration) without permission from the patent holder. But today there are tens of thousands of software patents and it is kind of impossible to keep an overview, especially for smaller companies. So the risk to get a punishment is very high, what could lead again to a monopoly.

Summarized, I think in such a fast industry like the software industry it is hard to find the perfect solution for patent durations. On the one hand, patents provide protection to those who invested a lot of money in innovations. This protection should be continued, but not for 20 years. On the other hand, patents can eliminate smaller companies from the competition, which leads to oligopoly or even monoploy.

In the post above, some arguments are given to keep actual patent length, to reduce it or to change the breadth and the application on software patents. I think the breadth would regulate by itself if we adapt the length of these patents. I will then give my opinion about the best length of the patents and explain why it will automatically change the breadth.

The first important point is that I exclude the possibility to delete patents because of the incentives it gives. In theory, without patents, the knowledge spreads out to all firms and it results in a bigger surplus for the consumers and the society. Unfortunately, without patents, there won’t be any incentives to innovate and then there won’t be any knowledge to spread.

There is now another question: if we have patents, what is the best length of these patents? Here I want to explain that if the patents are to long (unlimited or even 15-20 years), there will be little innovations with big profit for one company, nothing for the others and nothing for the consumer. The next big innovation will be after the patent limit, when the basis knowledge can be used to develop new softwares. We don’t want to wait 20 years between each improvement in softwares.

There is no innovation without past innovations. The length of the patents has to be the shortest possible because innovators need previous knowledge to create new softwares, which will give them profit during the patent and will be accessible to others firms to redevelop softwares after the patent.

Since there is no need of big fixed capital investment to develop innovative softwares or software firms, there is no need of long patents that allow to spread the amortization in many years. Furthermore, softwares are quick to enter the all market and rapidly generate profits.

The best length for the patent is then a length that interests developers to make innovation (the marginal dynamic gain of Nordhaus model) but that is as short as possible to decrease the marginal static loss. We have to calculate in how many years a software in general is profitable and we put the length to this number of years.
As far as I’m concerned, I would recommend a 3 years patent but I didn’t study the software market so I should follow my recommendations (search for the average length before profitability) before giving appropriate advice.

As I said in the beginning, I will now explain the natural regulation of breadth with shorter patents. Nowadays, some patents seems to be too broad because companies try to protect their idea and all derivations of this idea even if it doesn’t exist yet. With a 3 year patent, they won’t patent ideas that are not developed because they won’t be able to re-patent it when (or maybe “if” instead) they find derivations of their original software and it will fall into public knowledge before it is profitable. They will patent their idea and they have 3 years to derivate other patents before that others can use the first patent to derivate their own ideas.

I want to end up with one more recommendation, which is for politics but has an impact on the optimal patent length. There should exist an agreement on software patent length between every patent office for two reasons. First of all, softwares are easy to put in different markets all over the world. It is then easier to make rapid profits and the patent length may therefore decrease. If the product is imitate directly when it reaches new markets, firms won’t account it in their incentives to innovate. Secondly, due to globalisation and minimal need of fixed capital for softwares, knowledge can reach every innovator in the world and an agreement on patents would lead to better communications and more innovations as new technology reach everybody.