As someone who grew up imbibing the ideas of the likes of Richard Stallman and Lawrence Lessig, I’ve long thought that patent trolls were reserving a spot with the Walt Disney Company somewhere in the fourth or eighth circles of hell.
But, conversely, you get arguments (though it usually seems to come from IP lawyers themselves) about how patents are essential to innovation; without them, companies will be afraid to invest in bringing new products to market. Furthermore, in this “IP is wonderful” worldview, patent trolls – companies whose business model is “buy patents, sue companies that actually produce useful things with acquired patents” – help encourage innovation by acting as a market maker in patents, allowing inventors to profit and companies to license the IP they need.
So it’s interesting to read some actual evidence on the matter. Admittedly, it’s not been peer-reviewed yet, but it’s pretty striking. As the working paper by MIT academic Catherine Tucker explains, a patent troll called Acacia bought two patents that related to medical image retrieval software, and immediately commenced litigation against medical image software companies.
The lawsuit dragged on for a couple of years. In that time, the companies making medical imaging software didn’t release any new versions, and sales of their products cratered – and, overall, sales of all products in the category shrank by a third. However, smaller companies that were not sued, and companies that produced text-based medical retrieval software unaffected by patents, had continued sales growth.
So, in a nutshell, the patent trolls deprived the world of a couple of years of incremental improvement in medical imaging software, and delayed deployment of a technology which reduces medical costs and improves patient care.
That’s the economic story. The technical story is hinted at in the appendices to the working paper. While I lack the expertise to judge in a legal sense whether the patents were novel and non-obvious enough to meet the threshold for patentability, in my view they were nothing more than the obvious application of existing principles to a marginally-different problem domain.
In short, the patents describe a medical image retrieval system where the images are stored at different resolutions and compression levels, so low-resolution previews can be retrieved more quickly. Furthermore, in this system, textbook image processing techniques (such as edge detecting and contrast enhancement) can be applied. The second part is so obviously trivial it does not deserve further comment. The idea of low-resolution image previews, furthermore, is implemented in such methods as interlaced GIF, dating from 1987, and any image gallery software you’d care to name. Just adding “for a medical image retrieval system” is not in the least novel.
So, not only do we have an industry thrown into limbo for two years by a patent troll, we have them thrown into limbo by two joke patents that, under any sane version of patent law, should never have been granted.