The Incredible Shrinking Public Domain
On this site you can read about the legal changes that have eroded the public domain and find legal updates, including summaries of the recent cases that brought "Happy Birthday," "We Shall Overcome," and Sherlock Holmes into the public domain.
When Congress passed the first copyright law in 1790, the copyright term lasted for 14 years, with the option to renew for another 14 years if the copyright holder was still living. Before 1978, the copyright term was still 28 years from the date of publication, renewable once for another 28 years — but 85% of copyrights were not renewed and went immediately into the public domain. Under the 1976 Copyright Act, which went into effect in 1978, the term became 50 years from the date of the author’s death (with no need to renew to have the full term). And in 1998, the copyright term was increased to 70 years after the death of the author, and to 95 years after publication for corporate “works-for-hire”, locking up an entire generation of works for an additional 20 years.
The 1998 term extension — which increased the copyright term to life plus 70 years and 95 years for corporate authors — was not only granted to future works. It was retroactively applied to works that had already been created and enjoyed their full copyright term, and were about to enter the public domain. The public domain was frozen in time for over 20 years. In 2023, works published in 1928 - originally set to enter the public domain in 2003 - will finally go into our public domain.
In addition to lengthening the copyright term, recent laws eliminated the requirement that authors “opt in” to copyright protection by affixing a basic copyright notice — the word “copyright” or © with a name and year next to it. This change took effect in 1989, and copyright now adheres the moment a work is fixed, whether or not the author wants to protect or sell the work, with no easy way to "opt out" — think of all the amateur videos, travel photos, blog postings, witty musings, jam sessions, or useful tidbits that people want to freely share and don’t intend to commercialize. Estimates are that under the opt-in system perhaps only 10% of works included a copyright notice; the remaining 90% went immediately into the public domain. Compare that to today, when 100% of works automatically have life plus 70 years or 95 years of protection and are off limits to artists, educators, archivists, remixers, scholars, and everyone else who might want to freely use them.
The exceptionally long copyright term has also created a growing limbo of “orphan works.” These works are still presumably under copyright (only works published with authorization in 1928 or before are conclusively in the public domain), but the copyright owner cannot be identified or found (tracking down the copyright holders of older works is often impossible — read accounts of thwarted efforts to do so here). Orphan works comprise much of the record of 20th century culture — studies have found that only 2 percent of works between 55 and 75 years old continue to retain commercial value. For the other 98% of works, no one benefits from continued copyright protection, while the entire public loses the ability to adapt, transform, preserve, digitize, republish, and otherwise make new and valuable uses of these forgotten works. Read more about the current costs associated with orphan works here and here.
In the words of the US Copyright Office, “For good faith users, orphan works are a frustration, a liability risk, and a major cause of gridlock in the digital marketplace. . . This outcome is difficult if not impossible to reconcile with the objectives of the copyright system and may unduly restrict access to millions of works that might otherwise be available to the public.” So far, the US has failed to enact orphan works reform. However, in 2012, the European Union passed limited orphan works legislation – its Directive allows qualifying public institutions (such as libraries and museums) to digitize the orphan works in their collections and make them available to the public.
This steady erosion of the public domain is happening just as the Internet and digital technologies offer unprecedented opportunities to find, share, catalog, preserve, and remix its riches — foreclosing its enormous potential to feed creativity, innovation, democratic participation, and knowledge advancement.
Does all this mean that copyright is a bad system? Of course not. Copyright gives creators—authors, musicians, filmmakers, photographers—exclusive rights over their works for a limited time. This encourages creators to create and publishers to distribute—that’s a very good thing. But when the copyright ends, the work enters the public domain—to join the plays of Shakespeare, the music of Mozart, the books of Dickens—the material of our collective culture. That’s a good thing too! It’s the second part of the copyright bargain; the limited period of exclusive rights ends and the work enters the realm of free culture. Prices fall, new editions come out, songs can be sung, symphonies performed, movies displayed. Even better, people can legally build on what came before.
The basic principles of our copyright system are sound. But there is a consensus among academics, economists, and policymakers—including two heads of the United States Copyright Office—that the longer term is a “big mistake.” Why? Because its benefits are minuscule—economists (including five Nobel laureates) have shown that term extension does not spur additional creativity. At the same time, it causes enormous harm, locking away millions of older works that are no longer generating any revenue for the copyright holders. Films are literally disintegrating because preservationists can’t digitize them. The works of historians and journalists are incomplete. Artists find their cultural heritage off limits. Studies like the Hargreaves Review commissioned by the UK government, empirical comparisons of the availability of copyrighted works and public domain works and economic studies of the effects of copyright (other articles are here and here) all suggest that lengthy copyright extensions impose costs that far outweigh their benefits. In fact, economists who have modeled the ideal copyright term have uniformly suggested that it should be far shorter than it is right now. Some have suggested that it should be as short as 15 years. And every economic study has concluded that if there are to be copyright term extensions, they should not be retroactive.
So what can be done about all this? One obvious first step is legal reform that would give greater access to orphan works. The US Copyright Office has been engaged in efforts to find solutions to the orphan works problem (but, so far, none have been enacted). Authors and creators can also choose to license their work under more generous terms than standard copyright through Creative Commons licenses (for works like books, movies, music and art) or free and open source licenses for software. These open licenses create a privately constructed commons in which all can share freely. Fundamentally, though, the key is public education about the delicate balance between intellectual property and the public domain. That is the goal of our Center.
Legal Updates
“Happy Birthday” and key portions of “We Shall Overcome” are in the Public Domain! At the same time, countries are still lengthening copyright terms...
More Term Extensions Worldwide?! The global public domain is continuing to shrink. As explained above, there is a general consensus that the 20-year term extension in the US was a “big mistake.” Nevertheless, despite all of the evidence that it will do more harm than good, other countries are still agreeing to lengthen their copyright terms—not as a result of reasoned debate, but as part of compliance with trade deals. Canada and Japan both agreed to extend their copyright terms from life plus 50 to life plus 70 years as part of trade agreements (for Canada, it was the United States–Mexico–Canada Agreement or “USMCA”, and for Japan, the Japan-EU Economic Partnership Agreement). On December 30, 2022 Canada is adding 20 more years to its copyright term even though, quoting Professor Michael Geist, this would “cost Canadian education millions of dollars and would delay works entering the public domain for an entire generation.” In 2021, New Zealand joined the bad-policy party and agreed to extend its term as a concession to the UK in another trade deal, even though, quoting Michael Wolfe, it “would cost around $55m [NZ dollars] annually” without “any compelling evidence that it would provide a public benefit.” It has also agreed to term extension as part of an EU trade agreement. In the US, an earlier version of the Music Modernization Act would have granted extremely long terms to older sound recordings, giving recordings from 1923 a 144-year term(!). Thankfully, after pushback, these terms were shortened. Pre-1923 sound recordings entered the public domain in 2022 and recordings from 1923 are public domain in 2024.
“Happy Birthday” is in the public domain. In September 2015, a court invalidated Warner-Chappell Music’s claim to copyright in “Happy Birthday to You,” meaning that filmmakers, restaurants, and others no longer have to pay them licensing fees for using the song. (They had been collecting an estimated $2 million in annual fees.) Essentially, the court held that Warner-Chappell doesn’t own the rights to the “Happy Birthday” lyrics (the music is already in the public domain) because these rights were never properly acquired by its predecessor-in-interest. But the court stopped short of confirming that the lyrics are affirmatively in the public domain, because it did not need to resolve that larger question. (The available evidence suggests that the lyrics may have entered the public domain under several different theories — through publication without a proper notice, non-renewal of the copyright term, abandonment of the copyright, or lack of authorship.) In June 2016, the court approved a settlement declaring that the song is officially in the public domain. “Happy Birthday”’s copyright registration dated from 1935 — if it was valid, it would have lasted through 2030 because of successive term extensions.
“We Shall Overcome” is in the public domain. In September 2017, a court held that the melody and lyrics of the familiar verses “We shall overcome / We shall overcome / We shall overcome some day / Oh deep in my heart I do believe / We shall overcome some day” are in the public domain. These verses were based on an earlier public domain version of “We Shall Overcome” that was sung by tobacco workers on strike in South Carolina in the 1940s. The new version made only a few modifications, changing the words “will” to “shall” and “down” to “deep” and making minor melodic alterations. Because these changes were “too trivial” to meet copyright’s originality threshold, the court held that the verses in question were not entitled to copyright protection as a derivative work. The entities claiming copyright in “We Shall Overcome” were music publishers The Richmond Organization and Ludlow Music (not Pete Seeger). In January 2018, the parties settled, and the publishers agreed to dedicate the melody and full lyrics of the song to the public domain.
“This Land Is Your Land" lawsuit dismissed, unclear whether song is in the public domain. The Richmond Organization and Ludlow Music (not Woody Guthrie) are also the companies claiming copyright in “This Land Is Your Land.” Woody Guthrie’s opinion of copyright was clear from this notice on an early songbook:
“This song is Copyrighted in U.S., under Seal of Copyright # 154085, for a period of 28 years, and anybody caught singin it without our permission, will be mighty good friends of ourn, cause we don’t give a dern. Publish it. Write it. Sing it. Swing to it. Yodel it. We wrote it, that’s all we wanted to do.”
Richmond and Ludlow have a different approach — they famously claimed copyright infringement when JibJab Media used “This Land” in a parody of the 2004 Bush/Kerry presidential election. In 2016, the same law firm that prevailed in the “Happy Birthday” and “We Shall Overcome” cases filed a lawsuit seeking to establish that “This Land Is Your Land” is also in the public domain. According to the complaint, the song’s melody is in the public domain because it is almost identical to a Baptist gospel hymn variously titled “Fire Song,” “When the World’s on Fire,” “What you gonna do when the world’s on fire,” and “O My Loving Brother.” As for the lyrics, there is evidence showing that Richmond and Ludlow’s 1956 copyright registration is invalid because Woody Guthrie had already copyrighted the lyrics in 1945 when he published them with a copyright notice. Because he did not renew that copyright, it expired in 1973. (The lyrics were also published without a copyright notice on other occasions, and this could also invalidate Ludlow’s copyright claim.) In February 2020, a court dismissed this lawsuit, without ruling on the public domain question. Instead, the court decided that there was no longer a "live case or controversy" to adjudicate, because the copyright holders had granted a "covenant not to sue" the plaintiffs who were challenging its copyright.
These lawsuits demonstrate the complexity of determining the copyright status of older works, both in terms of highly technical copyright minutia and unearthing evidence from decades ago. To get a sense of the how difficult it is to establish whether works are in the public domain, you can read the “Happy Birthday” opinion here and an article about The Tangled History and Mysterious Legality of "Happy Birthday" here, read the “We Shall Overcome” ruling here, and read the “This Land Is Your Land” ruling here.
Sherlock Holmes and Dr. Watson are in the public domain!
In December 2013, a court held that the original Holmes and Watson characters are, in fact, free for all to build upon. (This decision was later affirmed by the appeals court.) These characters were first introduced in 1887, and featured in fifty stories that were published before 1923 and were therefore in the public domain, even under the 95-year copyright term. The court’s decision confirmed that all of the material within those stories is “free for public use.” The Conan Doyle estate had raised a novel legal theory that a “character is a work of authorship separate from the stories,” so that the copyright term for characters does not commence until “the creation of the characters [is] complete.” Because Holmes’ character development continued in a handful of post-1923 stories, his entire character—including all of the core aspects fully developed in the earlier stories—would therefore remain copyrighted. This bizarre theory would have created a special copyright term for characters, who could remain copyrighted long after the stories in which they appear enter the public domain, as long as their owners keep tweaking them in subsequent books. Not surprisingly, the court rejected this argument, finding that “It is a bedrock principle of copyright that once work enters the public domain it cannot be appropriated as private (intellectual) property, and even the most creative of legal theories cannot trump this tenet.” This loss did not stop the Conan Doyle estate from bringing yet another lawsuit in 2020, this time against Netflix for its Enola Holmes movie. The estate presented yet another strange legal theory: it argued that Holmes only began to "express emotion" and "respect women" in the later, still-copyrighted stories, and objected to Netflix's unlicensed use of those character traits. This argument did not succeed, and Netflix explained that personality traits such as warmth, empathy, kindness, and respect are unprotectable ideas under copyright law. The parties agreed to dismiss the lawsuit and settled in December 2020. As of 2023 the Doyle estate will no longer be able to make such baseless claims because the copyright will expire over Doyle's final Sherlock Holmes stories.
Google Books – The Digital Index of Alexandria – is LEGAL
This web page describes how expansions of the copyright term have eroded the public domain. Even as the public domain is shrinking, however, there are growing efforts underway to secure its promise of access our cultural heritage. One of these is Google Books.
In 2004, Google began an ambitious project—building a search engine for books. To do this, it could not simply use bots to crawl and index existing material. It needed the actual books—millions of them, hidden in the bowels of libraries. So it made a deal with libraries around the country, and started scanning their works. Many of these books were copyrighted. (The copyright term covers books published as far back as 95 years ago.) Google did not get permission to scan them because it would have made the project impossible. Of the millions of copyrighted books, most were out of print, and many did not have locatable rights holders. For those works, Google couldn’t get permission, because there was no one to ask.
You’ve probably used Google Books. You type in search terms, and it returns a list of books where they appear. If the book is in the public domain, you can see the whole book. If the copyright owner has a partnership agreement with Google, then you can see whatever pages the agreement allows. Potential copyright issues only stem from a third category—in-copyright books where there is no agreement. For those books, Google only shows three short snippets with your search terms highlighted. In addition, there is no advertising, and an “About the Book” page provides links to booksellers and libraries where the book can be obtained.
In 2005, Google was sued for copyright infringement. In November 2013, a court decided that Google Books was protected by “fair use”—the doctrine in copyright law that, broadly speaking, protects beneficial uses that do not displace copyright holders’ markets. The court found that Google’s use of copyrighted books was legal because it was “highly transformative”: it transformed the expressive text within books into something different, a new search tool. The court also emphasized the “significant public benefits” of Google Books. For the first time, people could find books that were “once buried in research library archives.” Beyond finding books, the “n-gram” project let researchers text-mine the full corpus of digitized books to track the evolution of words and phrases. Google Books allowed libraries to better serve their functions: they could use the digital scans to facilitate access for print-disabled patrons, and to preserve their holdings for perpetuity. Authors and publishers themselves benefited too: they gained new audiences and revenue streams from the people who were finally able to find their books. All of these benefits—widespread access, research, preservation, new markets – are just what copyright law is intended to promote. This decision was affirmed by the Second Circuit in October 2015.
Golan v. Holder—“Public” By Sufferance Alone (Or “We Can Take Stuff Away From You Too!”)
The Sherlock Holmes decision confirmed that material was in the public domain. However, the process can and has gone in the other direction. In the 2012 case Golan v. Holder, the Supreme Court held that Congress can remove works from the public domain without violating the Constitution. Even if the public now enjoys unfettered access to a work, Congress is allowed to take that work out of the public domain and create a new legal monopoly over it. What’s more, the Court declared, Congress can do so even when it is clear that the new right “does not encourage anyone to produce a single new work.”
If one reads Golan, one searches in vain for any limiting principle on Congress’s actions . . . Under the U.S. Constitution, says this case, the public domain is “public” only by sufferance. It may be privatized at any moment, at the whim of the Congress and without violating the Bill of Rights.
This decision marked a significant departure from the “bedrock principle” that once works enter the public domain, they remain there, free for anyone to use and build upon. The law at issue in Golan “restored” copyright to foreign works that had entered the American public domain for any of three reasons: the author failed to comply with copyright formalities, the U.S. did not have copyright relations with their country of origin at the time of publication, or they were sound recordings fixed before 1972. Some of these works were famous (including works by Picasso, Shostakovich, and Prokofiev), but many of them were obscure, older orphan works. In upholding the law, the Golan majority explicitly endorsed the position that the public has no legal rights to the public domain. While those who have copyrights enjoy vested, legally protected rights, “[a]nyone has free access to the public domain, but no one, after the copyright term has expired, acquires ownership rights in the once-protected works.” In a dissenting opinion, Justices Breyer and Alito asked “Does the [Constitution] empower Congress to enact a statute that withdraws works from the public domain, brings about higher prices and costs, and in doing so seriously restricts dissemination, particularly to those who need it for scholarly, educational, or cultural purposes – all without providing any additional incentive for the production of new material?” Their answer was “No.” (You can read more about this opinion here.)
NOTE: In recent years, the public domain has been diminished in many ways. Most relevant to Public Domain Day are the changes described above, but other expansions of intellectual property law have also contracted the public domain. These include changes in the subject matter covered by intellectual property and in the activities regulated. To learn more about these developments, see Professor James Boyle’s The Public Domain, available for free online here.
Public Domain Day 2024 by Jennifer Jenkins, Director of Duke Law School’s Center for the Study of the Public Domain, is licensed under Creative Commons Attribution 4.0 License.