Around the turn of the 21st century, both the public Human Genome Project and the private company Celera were actively sequencing genes. While the public effort’s results were free to all on the internet, Celera sequenced genes were only available on a data disc which was sold for millions of dollars. There was no government IP protection on gene sequences, so as soon as the Human Genome Project figured out a gene which Celera was selling, it immediately become free for all to use. A quirk in the nature of gene sequencing means that, to an approximation, the genes sequenced by Celera and the genes sequenced by HGP were random. Williams shows that Celera-sequenced genes, even 8 years later, still show less active research, and have fewer derivative products using their sequences. This implies that short-term IP may have large and persistent effects on downstream development. This has application to things like the Bayh-Dole Act, where lawyers convinced the US government to pass a law encouraging university labs to patent publicly-funded research, under the assumption that patented research will be more likely to move from the lab to the marketplace. Williams shows that this type of argument neglects the significant harmful effects on derivative products of such patenting.