Banning Sex Offenders from Social Networking Sites is Unconstitutional–Doe v. Jindal

By Eric Goldman

Doe v. Jindal, 2012 WL 540100 (M.D. La. Feb. 16, 2012)

Sex offenders–especially those who victimize children–are pariahs in our society. If it were possible, I bet many folks would favor blasting them off into space rather than “punishing” or “rehabilitating” them. Any legislative proposal to restrict the “rights” of sex offenders–even those who have served their time or otherwise been rehabilitated (whatever that means)–faces a one-sided political economy. No one ever sticks up for sex offenders, so laws targeting them typically pass quickly and non-contentiously.

In 2006, Congress passed the Adam Walsh Child Protection and Safety Act of 2006, which it extended with the Keeping the Internet Devoid of Sexual Predators Act (KIDS Act of 2008). (The names of these laws reflect common legislative tricks to speed passage and suppress opposition). Collectively, these two laws require sex offenders to submit their online aliases into a database which social networking sites can voluntarily access and then block the aliases if they choose.

Apparently unsatisfied with Congress’ efforts, Louisiana went a giant step further and prohibited certain sex offenders from accessing social networking sites, chat rooms or peer-to-peer networks. The punishment could be up to 10 years in prison with “hard labor” (this always reminds me of Cool Hand Luke). The law was defective on so many fronts, including:

* overly prophylatic. The law doesn’t criminalize misuse of a website; it criminalizes visiting the website due to the possibility that it might be misused. This is ridiculously overinclusive. It’s a bit like saying sex offenders can’t drive cars because they might drive to victimize children. Further, I’m not aware of any no social science validating the benefits of such a broad prophylactic ban on Internet technologies. Instead, hindering Internet usage by sex offenders deprives them of an essential tool for reintegrating into normal society. It’s like the laws that restrict sex offenders from living too close to schools; drawn too broadly, those laws ensure that the sex offenders have to move further away from their jobs, lie about their residence, or become homeless. At some point, eliminating the rights of sex offenders almost guarantees their further criminal behavior because they lack other meaningful choices.

* the law didn’t have a scienter requirement. A sex offender violates the law simply by visiting one of the verboten websites, even if the Internet user didn’t know that the website violated the law. For example, some websites provide a web interface to a BitTorrent P2P implementation, and the user may not even know. Or many websites have chat functions that are apparent only once you get there. The only sure way a sex offender could comply with the law is to avoid the Internet altogether.

* probation officers could grant permission to visit specific websites, but the law didn’t specify any standards for granting that permission. Furthermore, some sex offenders didn’t have probation officers (i.e., they had completed the probation), and the state law apparently directed those individuals to the federal court system–without providing funding or standards to the federal courts. You could read this opinion as the judge saying No Thank You!

While these drafting problems are serious, they are symptoms of an underlying root problem: social media exceptionalism. The statute’s real mistake is trying to carve out social media from the rest of the Internet and subject it to special treatment. As I’ve noted before, social media exceptionalism is bad policy, and it’s impossible as a matter of statutory drafting. See my 2009 article, The Third Wave of Internet Exceptionalism, and my 2007 summary, Social Networking Sites and the Law. Take a look at the specific statutory definitions here:

“Chat room” means any Internet website through which users have the ability to communicate via text and which allows messages to be visible to all other users or to a designated segment of all other users.

Huh? That describes every message board and tools for users to comment on blog posts and news articles. How about this definition:

“Social networking website” means an Internet website that has any of the following capabilities:

(a) Allows users to create web pages or profiles about themselves that are available to the general public or to any other users.

(b) Offers a mechanism for communication among users, such as a forum, chat room, electronic mail, or instant messaging.

Huh? That sounds like every UGC website, even if we read the connector between (a) and (b) as “and” and not “or.” Some examples: Wikipedia–yes. Boing Boing–yes. eBay–yes. Google–probably; it depends on how the statute contemplates a single company’s differently branded but integrated services.

The KIDS Act also had a definition of “social networking website”:

an Internet website– (i) that allows users, through the creation of web pages or profiles or by other means, to provide information about themselves that is available to the public or to other users; and (ii) that offers a mechanism for communication with other users where such users are likely to include a substantial number of minors; and (iii) whose primary purpose is to facilitate online social interactions

This is better that Louisiana’s definition, but not by much. Which of the sites I evaluated above aren’t covered by this definition? The harm that flows from this definition is much less than Louisiana’s criminalization with a decade in jail plus hard labor–the KIDS Act definition simply defines who can access the database of sex offenders’ online aliases for voluntary blocking purposes–but analytically it’s no more precise.

Naturally, it’s easy to pick apart the statutory language, but I can’t offer alternative language to fix the definitional overinclusiveness problem because I don’t think it’s fixable. None of these definitions come close to describing only the thing they target and nothing else. In my increasingly frequent talks about social media law, I make the point that social media law and Internet law are largely co-extensive because social media cannot be linguistically differentiated from the Internet ecosystem. This case shows that the overlaps are not only linguistic, but possibly Constitutionally required.

The ruling doesn’t require social media sites to allow sex offenders on their sites, and they can still use the Adam Walsh/KIDS Act database to block known sex offender aliases. (Of course sex offenders may not properly report all of their aliases, a general deficiency of the self-reporting database approach). Thus, striking down this law doesn’t immediately open up all of the Internet to the sex offenders. Nevertheless, it does mean that they can use the Internet without inadvertently committing a crime.