On 26 June 1997, in Reno v ACLU,[1] the US Supreme Court decided the fate of the Communications Decency Act (“CDA”), insofar as it criminalized the intentional transmission of "obscene or indecent" messages or information. In doing so, the Court made not only a finding that this provision of the CDA violated the 1st Amendment, but applied an approach to Internet cases with clear implications for cases the Court faces today.
Reno established that it is essential the Court recognize differences between the measured pace of judge-made law and the blistering pace of technology’s evolution, a point that is still cited by the Court today.[2] And, it identified that the capabilities and availability of the tools at issue have an important role to play in the constitutional analysis. As the Court continues to address Internet and technology-related constitutional cases, the importance of considering the capabilities of Internet tools may well be the most impactful legacy of Reno.
As Adam Thierer colorfully put it, “The Court concluded that there was “no basis for qualifying the level of First Amendment scrutiny that should be applied to this medium” and rejected the congressional effort to pigeonhole this exciting new medium into the archaic censorship regimes of the past.”[3]
It’s this last aspect of the decision – that the Court treated the Internet as its own medium – on which I focus here. In Reno, the Court carefully fashioned principles tailored to the unique characteristics of the Internet. It found the Internet was not comparable to newspapers, where ownership of papers of meaningful reach is economically concentrated. It was not comparable to broadcast radio and television because material online is sought out by intentional acts, rather than passively received.[4]
Reno indicates that even where particular rules were reasonable for earlier technology, the constitutionality of a given Internet-related rule must be considered anew in the context of the unique properties of networked information tools, as they currently stand before the court at the time of its ruling. And, as the District Court put it succinctly, “The Internet is… a unique and wholly new medium of worldwide human communication.”[5]
Which is to say, also, that implicit in Reno is the proposition that the constitutionality of a given law may well depend on the availability and capability of technology available at the moment – some rules that are constitutional may well become unconstitutional as technology advances, and vice-versa. And, as such, the Court should be cautious in setting long-term rules to govern a rapidly evolving medium.
These two points are made clear in both the main opinion by Justice Stevens and in the concurrence by Justice O’Connor (which Chief Justice Rehnquist joined). Justice Stevens cites two findings of the District Court: 1) that, at the time of trial, existing technology did not include any effective method for a sender to prevent minors from obtaining access to communications on the Internet without also denying access to adults;[6] 2) there are at least four “special attributes of Internet communication”[7] – all of which point towards far more open access to mass communications, and towards more democratic distribution of such tools.[8]
In Justice O’Connor’s concurrence, she observes that, as a matter of principle, cyberspace “undeniably” reflects a form of geography, on the assumption that websites exist at fixed "locations" on the Internet,[9] and that therefore zoning laws that restrict certain parts of the geography to adults could be constitutional. She agrees with the outcome here, though, on the basis that there was no technology available at the time to create such a “zoning” approach in a manner that does not preclude access to adults – and therefore does not sweep so broadly as to violate the 1st Amendment’s protections.[10]
As readers of this blog will likely surmise, I disagree with Justice O’Connor’s assumption that the Internet “undeniably” reflects a type of geography (aka “cyberspace.”).[11] As an illustration, what Reno referred to as "zoning" off Internet pornography isn't the same as "zoning" of construction and land use in the physical world. Rather, such Internet “zoning” was effectively a proposal to require Internet tools to recognize signals, e.g., code embedded in websites, that the content therein contained “indecent” material.[12]
In other words, it was a tool, rather than a configuration of physical space. It was useful to call it "zoning" because spatial metaphors are familiar, and because there was, for at least two Justices, a belief that “safe spaces for children” could comport with the 1st Amendment if technology could eventually make it possible.[13]
Indeed, today a variety of user-directed tools – which require no changes on the behavior of Internet publishers or access providers – are reasonably effective at blocking indecent (pornographic) content based on various heuristic and machine-learning approaches. And the availability of such tools was key to the Court finding a later Internet indecency law – the Children’s Online Protection Act (“COPA”) – unconstitutionally overbroad as well.[14]
This approach, assuming that constitutional outcomes may depend on the characteristics of technology, and may vary over time depending on changes in technology, is important today in other contexts – particularly 4th Amendment cases. 5 years ago, in U.S. v. Jones, the Supreme Court ruled that it was a 4th Amendment “search” to physically install a tracking device on the car of a suspect and to so monitor the suspect for a length of time.[15] While the Justices’s views varied on the reasons for that conclusion, Jones stands as a marked change from prior doctrine that one had no reasonable expectation of privacy in one’s public movements and activities.[16] Layered in among the reasoning for that change is the fact that technology has evolved so that far more information can be captured than when earlier doctrine was developed.
Jones differs from Reno, of course, in that one concerns surveillance rather than expression, but Jones is similar in that the question of whether a government action is constitutional depends on the way the technology operates, and to the extent to which it intrudes on constitutionally protected spheres. Relatedly, the Ashcroft case concerning the COPA statute (which followed from Reno) indicates the availability of tools is also important. And, as Kevin Bankston and Ashkan Soltani noted in analyzing Jones, a key factor is not only the availability of tools, but the relative cost of acquiring and using them.[17]
Now, in 2017, the court will revisit the “reasonable expectation of privacy” question in light of changes in technology when it considers, in Carpenter v. U.S., the use of a court order procedure to obtain business records revealing the location and movements of a cell phone user over the course of 127 days. The question there is whether the 4th Amendment instead requires the higher showing needed to obtain a search warrant. This case, too, will turn in part on the capabilities of the technology and, as in Reno, whether those capabilities allow the government to achieve legitimate interests through means that do not unduly intrude on constitutional freedoms.
In Carpenter, the petitioner argues that the “degree of invasiveness of the surveillance” involved distinguishes this case from prior precedent holding there is no reasonable expectation of privacy in records held by third-parties such as banks or phone companies.[18] The government, respondent, argues inter alia that changes in technology are not relevant, because a reasonable expectation of privacy cannot attach to 3rd party business records created by that 3rd party for its own purposes.[19]
But the government goes on to point out the use of a court order procedure is also constitutional because the technology used to develop the cell site location records at issue in Carpenter differs in precision from that in Jones (“as much as 12,500 times less accurate than the GPS data in Jones”), nor does it involve access to cell phone data which is as personal as that at issue in Riley.[20] In this way, the government, too, will contend that the capabilities and availability of the tools at issue are central to determining the lines of Constitutional protections.
Regardless of which view you find more persuasive, Carpenter is likely to also turn, in part, on discussion of the capabilities of the tools at issue, the extent to which those tools impact the constitutionally protected activity, and whether it is the appropriate time for the Court to set a long-term precedent in an area of quickly changing technology. And, we can consider the Reno decision a milestone highlighting the value of this approach to constitutional inquiry.
[1]521 US 844 (1997) https://www.oyez.org/cases/1996/96-511
[2]Packingham v. North Carolina 582 U.S. at ___ (June 19, 2017); online at /content/files/opinions/16pdf/15-1194_08l1.pdf (“The forces and directions of the Internet are so new, so protean, and so far reaching that courts must be conscious that what they say today might be obsolete tomorrow…”).
[3]Adam Thierer, “Celebrating 20 Years of Internet Free Speech & Free Exchange,” https://readplaintext.com/celebrating-20-years-of-internet-free-speech-free-exchange-8a10f236d0bd
[4]See, e.g., 521 US at 869.
[5]American Civil Liberties Union v. Reno, 929 F. Supp. 824 (E.D. Pa. 1996), at 844. Considering 1st Amendment issues through a medium-specific lens is not novel or unique, as the District Court noted. 929 F. Supp., at 873. What is important here, rather, is that both the District Court and the Supreme Court examined the specific characteristics of the Internet to fashion an approach tailored to that medium that was much more protective of free expression, given medium-specific capabilities such as the ability of users to select their own content, and the like. See n.7, infra.
[6]521 U.S. at 876.
[7]521 U.S. at 863. The characteristics noted by District Court Judge Dalzell were: 1) the Internet presents very low barriers to entry; 2) barriers to entry are identical for both speakers and listeners; 3) as a result of these low barriers, “astoundingly diverse content” is available on the Internet and 4) the Internet provides significant access to all who wish to speak in the medium, and even creates a relative parity among speakers.
[8]See 521 U.S. at 868 (noting “[t]he vast democratic forums of the Internet…”).
[9]521 U.S. at 890. Even here, she uses “locations” in quotation marks; signaling some uncertainty as to whether websites really are analogous to geographic points. As I’ve argued extensively in this blog, the geographic metaphor of “cyberspace” has manifest limitations; one is that a website does not in fact exist in a single “location”: several URLs may redirect to a given website’s content; the servers hosting a given website are likely geographically distributed and which server(s) deliver pages at any given time is a very dynamic process.
[10]521 U.S. at 890-892.
[11]Apparently, so do 3 Justices on the current Supreme Court; in Packingham Justice Alito’s concurrence (joined by Chief Justice Roberts and Justice Thomas) argued that “[c]yberspace is different from the physical world, and if it is true, as the Court believes, that “we cannot appreciate yet” the “full dimensions and vast potential” of “the Cyber Age,” ibid., we should proceed circumspectly, taking one step at a time.” See supra, n.2.
[12]See, e.g., Lessig & Resnick, “Zoning Speech on the Internet: A Legal and Technical Model,” (discussing a "kid’s zone browser." regime); online at http://bit.ly/2rOWqLZ. The technologies available at the time included the Platform for Internet Content Selection, wherein “tags” would be added to websites indicating a rating for that site (akin to how the MPAA rates movies); see https://www.w3.org/PICS/iacwcv2.htm. Challenges with this approach were many, however; among them were that many websites would need to adopt the labeling approach and that fashioning a legal or technical mandate to require such labels was challenging, especially given the bipartisan consensus for a “hands-off the Internet” regulatory approach. Also, as website technologies and business models evolved, content hosted by sites became more dynamic – varying both by what a publisher chose to host and as to what an interacting viewer might add or comment. Nonetheless, some parties did make a go of this access control approach. Microsoft created a “Content Advisor” tool, which read labels promoted by the Internet Content Rating Association, initially using the PICS format , and carried the tool in the Internet Explorer browser until its last supported version (Internet Explorer 11), although the feature was eventually moved largely out of sight due to underuse. See http://www.thewindowsclub.com/enable-content-advisor-internet-explorer-10-11
[13]Indeed, other attempts at “zoning” were tried, which would have relied not on access controls signaling the browsers of those seeking to access websites, but rather on a broad prohibition on using a particular Internet port to carry indecent content. A “port” is an indicator in communications between Internet-connected computers to identify a specific type of service (e.g., port 25 is generally associated with email, and port 80 with HTTP web surfing traffic. A proposal, backed by an entrepreneur and lawmakers in Utah, would have blocked obscene and indecent content from using port 80, with the goal of improving filtering capabilities. See http://www.deseretnews.com/article/660213162/Lawmakers-seek-ways-to-block-Net-porn.html
[14]COPA, enacted in response to the Reno decision, required all commercial distributors of "material harmful to minors" to restrict their sites from access by minors. It was found unconstitutional in large part because of the availability of “less restrictive means” in the form of filtering tools available to end users. The procedural history of the constitutional review of COPA is complex; for present purposes it’s sufficient to note that lower court findings striking down COPA were allowed to stand by the Supreme Court as the statute likely violated the 1st Amendment. See Ashcroft v. ACLU (03-218) 542 U.S. 656 (2004), 322 F.3d 240, affirmed and remanded; online at https://www.law.cornell.edu/supct/html/03-218.ZS.html
[15] United States v. Jones, 132 S. Ct. 945 (2012); online at: https://www.law.cornell.edu/supremecourt/text/10-1259. The Justices were, similarly, alternatively silent or unaligned as to the extent to which tracking must go (e.g., its duration) before a search warrant is required. This point too, is very likely to depend on the capabilities of the tools.
[16]See, e.g. U.S. v. Knotts, 460 U.S. 276, 282 (1983).
[17]“Tiny Constables and the Cost of Surveillance: Making Cents Out of United States v. Jones,” The Yale Law Journal Online 123:335 (2014); /content/files/pdf/1231_jjd1qz1e.pdf
[18]ACLU Petition for Certiorari (September 26, 2016) /content/files/wp-content/uploads/2016/10/16-402-cert-petition.pdf The ACLU also cites to Riley v. California, where the court relied in part on the capabilities and storage capacity of current cell phone technology to find that warrantless search of the contents of a cell phone incident to a lawful arrest violates the Fourth Amendment. Riley, 134 S. Ct. 2473 (2014).
[19]Brief for the United States in Opposition, /content/files/wp-content/uploads/2017/02/16-402-bio.pdf
[20]Id.