Page:Code Swaraj - Carl Malamud - Sam Pitroda.djvu/136

Code Swaraj entered search terms by agency, we were also able to pull articles by agency. For example, we had found 20,027 articles by employees of the Army Corps of Engineers and 45,301 articles from the National Institutes of Health.

For each of 29 major publishers, a statistically valid sample of articles was extracted, ranging from 50 articles for smaller publishers to 500 for the larger ones. The same process was conducted for each of 22 government agencies. All told, we pulled approximately 10,000 articles and performed a manual verification on each, looking for evidence of copyright assertions on the title page, examining the accuracy of our search results for false positives, and looking for indicators of the “officialness” of authorship, such as authors thanking their colleagues at work for their reviews or, conversely, indicating that the work had been conducted before they entered government service.

The results were pretty clear. Most of the articles we found were almost surely works of the U.S. government and in almost no cases was a proper disclaimer of copyright provided by the publisher. In most cases, the articles were carefully hidden behind a paywall and were certainly not available on the government’s web site and it was clear from an examination of the National Archive records disposition schedules for each agency that the Archives didn’t have a copy of these articles either.

The large-scale bibliographic search worked for most scholarly disciplines, but of course not for the legal profession which prides itself on willful ignorance of technology. The legal literature is, as a general rule, locked down so tight with exclusive vendors that it doesn’t make it into the general-purpose bibliographic search engines. However, I really wanted to know what the practice was in law journals, because this came down to a question of law. I asked law students around the country, led by one of my volunteers, Misha Guttentag from Yale Law School, to pull up some of the major law journals issue by issue and make spreadsheets with lists of articles that looked like they were by federal employees.

In addition to the university law reviews, another major powerhouse in legal publishing is the American Bar Association. I assigned that task to myself and manually examined several decades of articles across a few dozen different publications. I found 552 articles that sure looked like they were by federal employees, possibly in the course of their official duties.

An example was a Commissioner of the Federal Trade Commission briefing the antitrust bar on the agency’s regulatory actions and reforms for the upcoming