Are engineers really becoming less innovative?

Leland Teschler – Executive Editor
[email protected]
On Twitter @DW_LeeTeschler
Pundits have made headlines in recent years by claiming there is a slowdown in innovation. The evidence they usually cite comes from statistics about patents. The information firm Thomson Reuters, for example, recently used patent volume to conclude that the rate of innovation is slowing. The volume of global patents rose by 3.3% between 2013 and 2014, but this constitutes a slowdown in growth: In 2012, that rate was 20%; in 2013, 17.7%.

The implication is that slower growth in the number of patents means less new technology. However, patent information may be misleading. That’s one conclusion to be drawn from work by researchers from the Santa Fe Institute and Arizona State University. Writing in the Journal of the Royal Society Interface, they looked at U.S. patent records back to 1790 to determine whether the development of new technologies or combining existing technologies is more important in driving invention.

The researchers were helped along by the fact that the Patent Office classifies technologies responsible for inventive novelty through an elaborate system of technology codes. Examiners assign these codes to better summarize the claimed novelty. Patents can have not just single codes, but combinations of codes. And the codes can be grouped into classes of technologies. For example, the patent for the recombinant DNA technique is described as bringing together 24 distinct technologies of which 20 are drawn from the same class.

This classification scheme helped researchers get a handle on how technological innovation has evolved. During the first decades of the 1800s almost every patent represented entirely new technology, not combinations of existing technologies. One reason: The patenting system itself was an innovation, the researchers say, and inventors rushed to patent existing stuff they’d come up with over the years.

But if there has been a slow-down in invention, that slow-down happened in the late nineteenth century, at least if you go by the Patent Office’s technology codes. By the late 1800s about 150,000 technological functionalities had accumulated. Since then there have been few additions to the stock of individual codes, researchers say.

The conclusion from sizing up patents this way is that most patents are new combinations of existing technologies, and this has been the case for the past 200 years. In other words, the process of invention is driven almost entirely by combining existing technologies rather than by coming up with something completely new.

Put another way, most patents don’t describe radical new technology. And it has been that way for a long time. But that doesn’t mean inventors aren’t innovative.

Viewing patents from the standpoint of technology codes, fears of slowing innovation reduce to a simple question: Given that new technology codes don’t come along too often, how much further can we go by just combining existing technologies? The researchers say this basically boils down to calculating the theoretical bound for how many combinations of codes there can be. Run these numbers and you conclude the occasional introduction of a new code is more than enough to allow patentable innovation to continue for a long time.

But there is a rising chorus of commentators who say patents may impede rather than accelerate innovation. Typical of those in this camp is technology journalist Matt Ridley, who points out that the original idea of a patent was not to reward inventors with monopoly profits, but to encourage them to share their inventions. “Most patents are now as much about defending monopoly and deterring rivals as about sharing ideas. And that discourages innovation,” he says.

If Ridley is right, then, weirdly, we may argue that granting fewer patents, rather than more of them, is the way to boost innovation.