How Do Patent Laws Influence Innovation?
Evidence from Nineteenth Century World Fairs
by Petra Moser (MIT Sloan)
NBER Working Paper 9099
Full Text
Petra Moser’s paper examines the effects of patent laws across countries and asks whether countries with patent laws experience greater levels of innovation than those without them. Moser finds that strong patent-law systems do not necessarily increase the number of innovations, but they do affect the kinds of innovations a country produces.
Moser uses data from the catalogues of two nineteenth century world fairs: the Crystal Palace Exhibition, held in London in 1851, and the Centennial Exhibition, held in Philadelphia in 1876. Among the data are innovations that were patented and ones that were not, including innovations from three countries without patent laws: Switzerland, Denmark, and the Netherlands. At both exhibitions, a panel of industry experts evaluated all exhibits and awarded medals for the most useful and novel innovations. The awards offer a measure of the quality of the invention.
Theory
Moser begins with a number of theoretical predictions. Increases in the duration of patent grants, she argues, should increase the profit from innovation, and therefore increase the number of innovations a country produces. She cautions, however, that while the absence of patent laws lowers incentives for domestic invention, it facilitates domestic applications of inventions that were developed abroad. Scholars have argued that both Switzerland and the Netherlands benefited greatly from their unfettered ability to use and improve upon inventions patented in other countries.
Patent laws do not simply influence the number of innovations; they also help to determine the industries that see the highest levels of innovation. Moser argues that the relative usefulness of patent grants should be inversely related to an inventor’s ability to keep her innovations secret. The disclosure of an innovation required by a patent grant is very costly to an inventor whose innovation is difficult to imitate or reverse engineer. To the inventor whose innovation is relatively easy to reverse engineer, the protection of patent laws is essential for commercial success. Countries without patent laws, then, can be expected to exhibit higher proportions of innovations in industries where secrecy is effective.
Finally, the length of patent grants can also be expected to determine the distribution of innovations among industries. If patent life in a given country is short, then industries with higher initial profits from their innovations are more likely to patent. If the patent term increases, the advantages of early payoffs decrease relative to the long-term payoffs of the innovation, and industries with inventions that do not obsolesce quickly would most likely take advantage.
Empirical Results
In her analysis of the data Moser found that countries without patent laws produced as many innovations as countries with patent laws when adjusted for population. The type of innovations produced, however, differed between countries with and without patents, and among countries with different patent lengths. As it turns out, patent laws seem to influence the industrial composition of innovation, but they do not necessarily increase innovation per se.
Countries without patents were particularly inventive in the fields of food processing and the manufacture of scientific instruments; indeed, Denmark and Switzerland, both without patent laws, had the largest share of innovations of any country in scientific instruments. However, the patentless countries were not very innovative in machinery fields, especially large-scale manufacturing, and in mining and mineral products.
Countries with short patent grants (under 10 years) were more similar to countries without patent laws than to those with long patent grants. They exhibited significantly higher proportions of inventions in food processing than those countries with longer patent grants. Moser explains that the absence of patent laws guided innovative activity toward areas where patent grants were not necessary to protect innovations. She cites historical evidence that secrecy, rather than patenting, was widely used to protect innovations in scientific instruments and food processing.
Overall, introducing strong and effective patent laws to countries without any such institutions, Moser concludes, would have a greater effect on the direction, rather than the rate, of innovation. In other words, patent laws do affect innovation in industries, but the effects are in kind, not in number.
