Anjali S. Dalal / Administrative Constitutionalism and the Growth of the Surveillance State
People apparently think it’s worse for the gov’t to shut down Reddit in response to IP claims than for the government to pervasively surveil for national security purposes. Is this really justified, or merely something we haven’t thought through?
Cellphone carriers received 1.3 million demands for subscriber information for gov’t investigations. Law enforcement requests to third parties have been growing 12-16%/year. Twitter received 849 requests globally, 679 from the US; subpoena is sufficient, without a warrant. Third party doctrine eviscerates any privacy interests in information you share with third parties. This may have made sense at one time, but we now have no idea who has access to our info in any internet interaction—so many different intermediaries touching the information. Little understanding of who has possession and control.
CISPA: allow unprecedented sharing between private companies and the government; creates exceptions to all privacy laws; undefined national security purposes; no privacy minimization required; entities are granted full immunity for any lawsuits; no FOIA available. But in addition to CISPA, there are a slew of other laws & practices that allow gov’t surveillance.
Emerges when we elevate national security over civil liberties; the FBI overtly acknowledges this in its priorities. Our government feels it’s doing the right thing by engaging in pervasive surveillance. NYPD targets Muslim student groups, framed as guarding against the threat of terrorism. Bloomberg: “this is what you want the NYPD to do.” That’s the concern with surveillance culture: we’re proud of it.
Is this level of gov’t surveillance new? No and yes.
Administrative constitutionalism uses agencies as frontline of constitutional interpretation. Bill Eskridge & John Ferejohn moves beyond descriptive account and argues that agencies are and should be constitutional norm entrepreneurs—trial balloons. Consider the DoJ’s norm entrepreneurship. Church Committee in 70s investigated US surveillance practices and recommended mostly speech-based ways to reform the FBI. Other guidelines for intelligence gathering exist—Levi Guidelines. By their nature, intelligence investigations chill speech, so you need deliberation before such investigations. Slowly, over 40 years, we relaxed the restrictions, allowing more and more investigation. The probability of stopping a crime needed to be lower. The kinds of crime covered became broader and less associated with national security. Freeing field agents from the bureaucracy that hindered them. By 2008, can open investigations for “assessment” with no factual predicate.
Also, things are more secret now: citizens don’t know what’s going on. FBI conducted 82,000 assessments on individuals, but the process of having to open a file (which had no factual predicate) was too burdensome, so they created a pre-assessment procedure which was even easier and less documented.
Her argument: administrative constitutionalism has slowly shifted the balance between national security and free speech. This has allowed the surveillance culture to become entrenched.
Q: Connect the story you told to the takeaway—this is about norm entrepreneurs. Playing devils’ advocate: you’ve shown that a bad thing happened, but should we blame administrative constitutionalism/norm entrepreneurship? Doesn’t it depend on other things—there may have been popular control all along; Congress had some ability to check in and following 9/11 Congress was supportive—not just the FBI over/reacted. Does administrative entrepreneurship work better with other agencies? How generalizable is this?
A: There wasn’t actually any control over the agencies; there wasn’t anything explicit. In 2008, Mukasey finally went to Congress to talk about the guidelines—not to get approval—and a Senator even thanked him! This might be implicit consent, but there was no outward check—no oversight by Congress, judiciary, public.
Q: why is agency singular? When things go off the rails is when you only have one agency, whereas checks and balances can be created by agencies fighting/turf battles. CALEA: FBI wants more authority and Commerce is resisting.
A: Definitely. But civil liberties lacks a constituency.
Q: Georgetown Big Data event: General Counsel for NID argued that it was the agencies making these decisions; when they want to share info, they enter into complicated negotiations about who will control the info, where it will be housed, etc. Memoranda of understanding are where privacy rules come. But that’s a ridiculous way to replace the 4th Amendment and public oversight. Demonstrates that agencies are convincing themselves that they are making up the rules and taking care of us through these less observable and more informal processes.
Q: why should the third party doctrine care about whether the discloser knows the identity of the third parties or their number?
A: if the doctrine is based on the idea that the disclosure is making an active, knowing, relational decision, then the predicate is gone—I can’t hold these entities responsible for a wrongful disclosure and there’s no way for an individual to make a rational decision about disclosing to all these.
Felix Wu / The Ontology of Speech
Questions of the form “is X speech?” Why do we ask this question? The way in which the question is framed isn’t a helpful starting point, if that means “is X capable of serving an expressive purpose/conveying a message?” Anything has that capability, including a speeding car. Teasing apart what’s really going on: There’s a need to balance the expressive purposes but also the non-expressive effects of this act or activity in the world. We can’t have either inquiry alone. Resolving the conflict is hard and doesn’t claim to have a complete answer. Because often we ignore one of the two strands in the caselaw. Worthwhile to discuss even if we don’t change our mind about how cases come out.
Snyder v. Phelps. Breyer: we never let people slug other people in order to get a message out. What’s going on in Snyder? The protestors are trying to convey a particular political message. But the protest itself might well be understood as directly inflicting emotional harm at a funeral. Agreement/disagreement with the political message isn’t the source of harm—or is it? That kind of harm results from expressive purpose rather than nonexpressive effect. SCt makes prophylactic rule: too hard to tell difference between emotional harm from the means of conveying the message and the emotional harm from the message, so we give breathing room to speech. So in which situations do we need prophylactic rules?
So, start with whether search results are speech? Search results = Google conveying a message about what results you’d like to see? In one sense, yes, a message is being conveyed. It is also limiting your ability to reach other things or directing you to reach some things over others, like a retailer deciding what to stock on shelves—though 1A scrutiny would never apply to the latter, even though one could claim that the retailer’s selection sends a message. By speaking in those terms, we can start to think about what weight to grant to Google’s interest in this case.
What about data collection? The effects of the sense of being surveilled might itself be a kind of independent harm that could be cognizable, but that’s also relevant with respect to information dissemination. The harm of feeling watched also occurs when information is used. So it’s not clear that collection and dissemination are good lines for 1A purposes. That doesn’t mean the 1A necessarily requires scrutiny for both of these; rather it might mean neither. Harder to draw a line between those two forms of privacy harms.
Copyright: traditionally the view has been that copyright has been justified because it encourages more speech. His approach: copyrighted works have expressive potential and nonexpressive effects. Dissemination is a way of monetizing an intangible consumptive good. Consumption isn’t all there is, but there are two aspects. We lose sight if we only focus on the ability to speak through copying/creating transformative works; look at the nonexpressive effects—the inability to monetize consumptive goods. Thinking about it in those terms might change the 1A valence of some aspects—basic rule against copying v. idea/expression or fair use. Might distinguish between substantially identical copies and something else, especially w/r/t intermediary liability. Could impose obligation on ISPs to reasonably go after substantially identical copies, but CDA immunity for everything else.
Fred von Lohmann: the first case recognizing code as speech was Bernstein v. DoJ—a licensing requirement imposed on cryptography related communications. What does your analysis have to say about prior restraint?
A: Not sure—code is a clear example of this dual situation. Traditionally the bad effects are the result of the expression; if the relevant effects are nonexpressive rather than expressive it’s different.
Carlos Luna: is this an issue of the policies behind why we have free speech or a disconnect between cases?
A: the sense in which I use “ontology” is questioning whether speech exists as a thing. The thing we need to protect is not a category of things but a purpose to which things can be put. (Paging Mike Madison.)
Luna: then shouldn’t the ultimate question be does the construct itself meet the policy that underlies the category “free speech”?
A: yes, then you need a theory of the 1A. Wants to carve out nonexpressive effects, things that aren’t the result of the expression in the speech act.
Blitz: would help to distinguish your proposal from a requirement of content neutrality. Some people might respond: what Corley does is says that we apply O’Brien for a content neutral regulation. You’re using expressive not just to cover content-based but also maybe some content-neutral categories. Is it only communicative or are there other ways to infringe on the expressive aspect of an activity?
A: Some of what I mean is content-based under current doctrine. If we treat data/code/search results differently—that might be understood as content-based. It’s not viewpoint discriminatory but it is content-based. Blitz: but that might still be nonexpressive. Wu: Google is expressing something, but there’s also a nonexpressive aspect. That could be classed as content based but he’d still ask about which strand is at issue. He’s trying to define a class of things not even subject to intermediate scrutiny—that’s how content neutral it is.
Jane Bambauer: Not a clear enough line between nonexpressive and expressive harm. Is the only form of expressive harm is the harm I feel from disagreement? That seems troublingly narrow. The emotional distress of a family is not nonexpressive in her view. NYC’s decision to throw up billboards to shame teen mothers—where would you put shame? Is that an expressive harm?
A: Disagreement is not the only form of expressive harm. It’s expressive if it has the potential to change your mind, not your viewpoint, but change how you think about something. Trying to distinguish that from emotional hurt. More along the lines of an insult—the exclusion for fighting words. It doesn’t really change your mind about anything, just designed to convey or induce an emotional state. The father in Snyder wasn’t going to change his mind about anything; his emotional distress could’ve been the result of the expression rather than the way in which the protestor was doing it.
RT: Fred Schauer and what is thinkable—you are starting with the edge cases; all 1A claims involve claims that nonexpressive harm is happening, otherwise they wouldn’t be suppressing it and there’d be no rational basis for the regulation. Why doesn’t draft card burning have the potential to change your mind? Why don’t high drug prices have the potential to change your mind about whether gov’t regulation of the healthcare market is a good idea? Doesn’t bullying try to make you hate yourself? Or perhaps try to make you take a swing? Copyright: of course the harm comes because of the expression: because I torrented and enjoyed The Avengers, I don’t pay for it! [NB: purely hypothetical. I paid for and plan to enjoy The Avengers many times!] An emotional/rational distinction also won’t work because we only think with our emotions.
A: is the bully conveying anything? Yes. There’s no such thing as a purely nonexpressive activity. My claim is that’s true of anything. [And is there such a thing as a purely expressive activity?]
Defamation’s harm is exactly that someone else’s mind has been changed. Economic harm of copyright infringement comes from consumption. Not claiming it’s physically possible to have one without the other, but that they are conceptually independent.
[RT: But then they are always conceptually independent; this is true in defamation too where we can just have the concept of economic and emotional injury from the defamation, just as the alleged economic harm of copyright infringement works through having consumed the infringing material. I take it Wu’s argument is that you’re only harmed by defamation if someone’s mind is changed, but the copyright owner is only harmed by infringement if someone was exposed to and satisfied by—satisfaction being a result of mental processes—the work and therefore didn’t buy it (or didn’t like the work and would’ve bought the work before finding that out absent infringement).]
Marc Jonathan Blitz / School, Work, Speech, and CyberKant
Speech as designed to safeguard a certain sphere of autonomy and activities related to it—Rodney Smolla: speech is connected to thought in a manner that other forms of gratification aren’t. Steven Heyman: Boundary between the outward realm of the state and the inward life of the individual. Social convention, historical accident, and “satisficing” constitute First Amendment law—even if you can’t protect all activities key to autonomy, you protect enough to preserve the citizen against the state. The assumption is that we shouldn’t do this retail but rather wholesale: very strong defaults even if certain things don’t fit well into abstract definitions of expression.
Public forum doctrine began as creative move by the courts. Kant asked how in a free society we can be so limited in professional, military, and other ways? Can be free as a scholar while constrained as a priest—you can step out of your role and that defines freedom. Robert Post uses this to distinguish between democratic legitimation and democratic competence. The reason he focuses on speech and work is that Kant’s question remains important. A priest can’t just step out of his role and criticize everything he stands for and still expect to give Mass the next day. Same with schools and workplace speech. The direction of free speech law is shortchanging the necessary distinctions by pushing the misleading view that speech not about political topics or topics of public concern gets very little protection, sometimes none. This conflicts with speech as scaffold for mental autonomy.
Internet connection: the worry is that the internet is everywhere and can affect everything, and on that basis people will lose their jobs and get kicked out of school.
Bambauer: Liked the idea of promoting 1A approach satisficing in the sense that we build up categories (whether rules or standards) to know what the safe spaces are.
A: Yes, but if we allow room for history, the problem is that it may reduce to power politics: in 1939 the Court saw an opportunity in Hague v. CIO to make parks safe for speech. 1A institutionalism might provide a different way to draw boundaries, other than public forum doctrine—identifying spaces for preserving speech (institutions) but admitting that there’s always going to be some arbitrariness.
Anupam Chander: what do you think is your most controversial point?
A: maybe not controversial enough. Might not have much company in liking public forum doctrine. Kennedy has defended it v. Breyer’s balancing. Kennedy seems to like the rigor of categories.