What determines the strength of a patent

Research conducted at Cass Business School has shown that the number of claims and classes in a patent can be informative as to the amount of protection that patent provides.

A patent is defined as a "government authority or licence conferring a right or title for a set period, especially the sole right to exclude others from making, using, or selling an invention". Millions of patents are filed each year by firms all over the world for inventions that they jealously guard. Just how effective though are patents in protecting the inventing firms from other companies active in the same area? Research conducted at Cass Business School has shown that the number of patent claims and classes in a patent can be informative as to the amount of protection it provides.

"Imagine a space that holds all the ideas that have already been created, as well as all those that have yet to be generated. We can think of a specific invention as occupying an area in that space. A patent is essentially the temporary right to exclude others from making, using or selling an invention positioned in that area in exchange for its disclosure", says Elena Novelli, author of a paper published this year in Research Policy, one of the top international academic journal in technology studies. As the paper explains, a patent would have limited value if it did not protect an inventing firm against small modifications to their original idea. Hence, the patent system allows the inventing firm to specify in the patent claims any possible improvements or variations that could be made to the patented invention to adapt it for different uses. To illustrate this, consider a patent for a mobile phone. The device consists of a main body with an input device, such as a keypad. To grant the invention broader protection from the patent the inventor could specify that variations would still be considered part of the original invention - a variation such as a touchpad instead of a keypad.

Claims may refer to marginal variations to the invention (e.g. the diameter of a component) or to more 'distant' variations i.e. they may refer to completely different materials from which the same component could be made, and could be made to adapt the invention to multiple uses. They can even refer to different technological domains - for instance textile, metal working, liquid crystals. In the US patent system, the empirical context of this paper, this is captured in the classification associated to the patent.

Novelli's paper shows that it is not enough to look at the number of claims when determining the extent of protection afforded by a patent. Rather, the way in which those claims are positioned in the inventive space, i.e. whether they refer to variations so different that they span across technological domains instead of being concentrated in a few ones, plays a fundamental role. "A higher number of claims act as a deterrent to other firms from building on the knowledge underlying the focal patent, because there might be a higher risk of infringement" explains Novelli. However if the claims are too dispersed the protection they provide becomes sparse for many reasons: "First, it is less likely that the focal firm has the capabilities to develop the patent along so many directions; at the same time, patents classified in more different classes will also have greater visibility, and be more likely to 'cross' more firms' search processes. The problem is complicated by the fact that in such a case it is more difficult for the inventing firm to monitor a broader area to detect possible infringements".

The paper shows empirically that patents with a high number of claims tend to be the basis for subsequent patents by the same inventing firm. However this is not the case when they refer to very distant variations to the invention: other firms (but not the inventing firm itself) tend to generate patents that build on the firm's original invention. This is an interesting finding, as existing research in this area had commonly assumed that the number of classes in a patent could be a proxy for its scope.

A possible implication from this research therefore is that patents grant higher protection when their claims are denser, i.e. when the number of claims per number of classes is high. But what enables firms to identify a greater number of patent claims, and what determines their positioning across a greater number of technological domains? Surprisingly, there has been limited research exploring this question. Dr. Novelli's paper identifies firms' prior investments in scientific knowledge and in related inventive experience as two factors affecting patent scope.

By enabling firms to abstract from the specific context of the invention, scientific knowledge facilitates the identification of variations to the initial invention, even very distant ones, and these variations can be included in the patent claims. "For instance, the invention of photonic-crystal fibres - a new class of optical fibres - was developed through the transfer of knowledge from the principles of quantum mechanics to the field of optics, and built on the theoretical idea that light could be trapped in photonics 'bandgaps' in a similar way to how electrons can be trapped in the energy gaps of a lattice" explains the author. Having prior related inventive experience can also help the search for variations to the original invention, increasing the number of claimed variations the inventing firm can incorporate in its patent documents. However, having much relatedness prior experience is likely to reduce the extent to which the variations identified span multiple domains.

Dr Novelli has developed this paper in the context of her ESRC Future Research Leaders grant. It is available to read in full at ScienceDirect.