Is there any evidence of chilling effect on reading/web searching behaviour from fear of being watched

In a privacy context, the term “chilling effect” refers to the way in which people could modify, change their behaviour, or self-censor because of the fear that they are being watched, or they know for certain that they are. The term can be applied in a number of different areas of law. For example, in the law of libel, somebody might choose to exclude a particular paragraph or sentence from an article or indeed might opt not to write the piece at all for fear of falling foul of libel law.

Jeremy Bentham designed a prison as a means of keeping prisoners under control consisting of cells which were arranged radially in a circle and a central tower. Known as the panopticon, the prison’s design consisted of cells that were backlit so that anyone in the tower could see anything that was going on in the cells, but that crucially the prisoners themselves weren’t able to see whether anyone was in the tower. The idea behind the building’s design was that prisoners would assume that the guards were watching them, and that they would act accordingly. The panopticon is therefore an instrument for exerting power and control.

Michel Foucault used the panopticon as a metaphor for the impact of surveillance where you feel visible and exposed, and feel the power of the potential surveillance (Foucault 1977).

“Records must be protected from the self-appointed guardians of public and private morality and from officials who might overreach their constitutional prerogatives. Without such protection, there would be a chilling effect on our library users as inquiring minds turn away from exploring varied avenues of thought because they fear the potentiality of others knowing their reading history” (Quad/Graphics v. S Adirondack 174 Misc.2d 291 (1997), 664 N.Y.S 2d 225. 1997).

You might not visit certain websites if you thought that the government would use this as evidence against you. You might choose to limit your reading choices to the mainstream, to the bland and the boring rather than read something controversial, racy, or embarrassing.

Is there any evidence that library users change their behaviour when they think or know that they are being watched?

Some of the examples below are specifically in a library context. Given that many library visits aren’t to borrow books, but to use the computers, the other examples are also relevant as they cover searching the internet:

  • A study at Central Michigan University’s Park Library found that LGBT material was borrowed 20% more if done by self-check than using the traditional circulation desk (Mathson, Hancks 2008).
  • Following the Snowden revelations of mass surveillance, users search behaviour changed, including a drop in traffic for search terms rated as being personally sensitive (Marthews, Tucker 2015).
  • In Norway you can look up anybody’s tax records. However, people must now log onto the tax system and thereby leave a trail of which records they have looked at. The number of searches fell by 90 percent from 2013 to 2014 as a result.
  • A young woman stopped short of printing out her research on sexually transmitted diseases when she learnt that the printer was at the front desk. Source: “Privacy concern, printer control clash at library” (Fillo 1999).

The right to read

Neil Richards argues for a certain kind of privacy, one which is different from tort privacy, which he refers to as “intellectual privacy”, and which he believes to be essential if we care about freedom of expression (Richards 2015). Indeed, he says that privacy and (freedom of) speech are not conflicting values but are mutually supportive. The three most important elements of intellectual privacy that Richards identifies are:

  1. the freedom of thought;
  2. the right to read (and engage in intellectual exploration),
  3. and the right to communicate in confidence

Reading is often an act of fantasy, and fantasy cannot be made criminal without imperiling the freedom to think as one wants. Moreover, the chilling effect of such an intrusion into intellectual privacy could cause others to skew their reading habits for fear of attracting the attention of the government (Richards 2008)

The United States Supreme Court first protected the right to read anonymously in Lamont v Postmaster General 381 U.S. 301 (1965), in a decision that struck down as unconstitutional a law requiring individuals to identify themselves in order to receive publications that were allegedly Communist propaganda. The Court’s opinion relied on the “chilling effect”.

Mass surveillance by the state results in the chilling effect of limiting our intellectual privacy, inhibiting our ability to seek out new ideas that conflict with the status quo and therefore inhibiting our autonomy (Clark 2016)


Libraries are not the only place that private records of intellectual activity can be found. Other examples include:

  • lists of who purchased which books from bookstores; or
  • publishers’ records of who subscribes to a particular journal title.

In re Grand Jury Subpoena to Kramerbooks & Afterwords, Inc., 26 Med. L. Rptr. 1599 (1998) the U.S. District Court for the District of Columbia held that the government must meet strict scrutiny when defending subpoenas to a bookstore for customers’ book-purchase records. The case involved the Independent Prosecutor Kenneth Starr’s attempt to obtain Monica Lewinsky’s book-purchase records. Starr was trying to establish whether or not Lewinsky had purchased a novel by Nicholson Baker entitled “Vox” which chronicles an intimate and graphically sexual telephone conversation. In holding that the government would have to overcome strict scrutiny, the court stated that “[t]he bookstores and Ms. Lewinsky have persuasively alleged a chilling effect on their First Amendment rights.”

In Tattered Cover Inc v City of Thornton 44 P. 3d 1044 (Colo 2002) six police officers entered the Tattered Cover Bookstore in Denver with a search warrant for a specific customer’s book-purchase records. Joyce Meskis, who was the store’s owner, refused to hand over the records because she was concerned that complying with the warrant would violate the customer’s First Amendment and privacy rights.

In holding that Tattered Cover did not have to turn over the customer’s book records, the court first explained that its decision vindicated both the rights of the bookstore and of the book-buying public in general. Both the United States Constitution and the Colorado Constitution safeguard the right of the public to buy and read books anonymously, free from governmental intrusion.

The court stated that “when a person buys a book at a bookstore, he engages in activity protected by the First Amendment because he is exercising his right to read and receive ideas and information.” Otherwise, the right to free expression would be meaningless if the right to receive the thoughts that someone else was free to express was not similarly protected. But the court went further, expressing the right in more affirmative terms: “Everyone must be permitted to discover and consider the full range of expression and ideas available in our “marketplace of ideas”. Additionally, fearful that any governmental inquiry into the book buying habits of the public would “almost certainly chill their constitutionally protected rights,” the court emphasized the importance of anonymity.

Accurate information can tell inaccurate stories:

– reading a murder mystery does not make someone a murderer

– tracking someone’s location to having been outside a shop minutes after it has been robbed doesn’t make them the robber

– reading a book about aphasia doesn’t automatically mean that the person reading the book is afflicted by the condition

People read books for an infinite variety of reasons, and drawing generalized conclusions from another’s reading choices wrongly assumes that the most obvious reason is always the correct one.

Clifford Lynch (2017) says “remember that knowledge of actual reading activity rather than simply knowing what texts have been accessed or acquired still does not guarantee understanding of the values, beliefs, opinions, or intentions within a given human mind. We can only hope that governments, and commercial data collectors and exploiters, know this as well” (Lynch 2017)

Why does this matter? It matters because of the way in which courts have used the records of what people have read as an indication of intent (see, for example, United States v Curtin 489 FJd 935, 956 (9th Cir. 2007))

Penney (2016) undertook an empirical study providing evidence of regulatory “chilling effects” of Wikipedia users associated with online government surveillance. The study explored how traffic to Wikipedia articles on topics that raise privacy concerns for Wikipedia users decreased after the widespread publicity about NSA/PRISM surveillance revelations in June 2013. Using an interdisciplinary research design, the study tested the hypothesis, based on chilling effects theory, that traffic to privacy-sensitive Wikipedia articles reduced after the mass surveillance revelations. Penney found not only a statistically significant immediate decline in traffic for these Wikipedia articles after June 2013, but also a change in the overall secular trend in the view count traffic, suggesting not only immediate but also long-term chilling effects resulting from the NSA/PRISM online surveillance revelations. This study is among the first to evidence—using either Wikipedia data or web traffic data more generally—how government surveillance and similar actions may impact online activities, including access to information and knowledge online. (Penney 2016)

Proving a causal link to the “chilling effect” is not easy. As Penney acknowledges: “Privacy theorists, security researchers, and social scientists have also expressed skepticism about the possibility of large scale chilling effects caused by online surveillance. One reason for such skepticism is increasing public acceptance of, or desensitization to, privacy and surveillance concerns, particularly in new technological contexts. Indeed, some research in the field suggests that any chilling effects would, at the very most, be temporary or ephemeral, as online users have changed their behavior in response to shifting norms” (Penney 2016)

Social science research has long illustrated that self-reported or expressed concerns about privacy do not necessarily reflect people’s actual behavior online, a phenomenon sometimes referred to as the “privacy paradox.”

The theory of “chilling effects” received a comprehensive exploration in Schauer’s Fear, Risk, and the First Amendment: Unraveling the “Chilling Effects Doctrine. Schauer conceived of chilling effects as primarily resulting from people’s fear of prosecution or legal sanction and the uncertainties of the legal process (Schauer 1978)

(Marthews, Tucker 2015) found a statistically significant 5% reduction in Google searchers for certain privacy-sensitive search terms after Edward Snowden’s revelations in June 2013. Their study not only provides evidence of chilling effects, but also offers a research design that may be employed to study chilling effects in other online contexts (Penney 2016). However, as Penney says “the authors obtained their data from Google Trends, which provides Google search data in “normalized” or adjusted format. The search data is normalized in two ways. First, the data represents only a percentage of total Google searches for any given term. Second, Google then “adjusts” the search data to render comparisons across regions more easily; these results are further “scaled to a range of 0 to 100”. This, the authors admitted, meant it was “harder to make projections” based on the findings of the study— such as resulting “economic outcomes”.

PEN America’s research documents the chilling effect of encroaching surveillance on creativity and free expression:

  • 28% have curtailed or avoided social media activities, and another 12% have seriously considered doing so;
  • 24% have deliberately avoided certain topics in phone or email conversations, and another 9% have seriously considered it;
  • 16% have avoided writing or speaking about a particular topic, and another 11% have seriously considered it;
  • 16% have refrained from conducting Internet searches or visiting websites on topics that may be considered controversial or suspicious, and another 12% have seriously considered it;
  • 13% have taken extra steps to disguise or cover their digital footprints, and another 11% have seriously considered it;
  • 3% have declined opportunities to meet (in person, or electronically) people who might be deemed security threats by the government, and another 4% have seriously considered it. (PEN America 2013)

Stoycheff undertook a study exploring how perceptions and justification of surveillance practices may create a chilling effect on democratic discourse by stifling the expression of minority political views. Using a spiral of silence theoretical framework, knowing one is subject to surveillance and accepting such surveillance as necessary were found to act as moderating agents in the relationship between one’s perceived climate of opinion and willingness to voice opinions online. (Stoycheff 2016)


CLARK, I., 2016. The digital divide in the post-Snowden era. Journal of Radical Librarianship, 2, pp. 1-32.

FILLO, M., 1999. Privacy concern, printer control clash at library. The Hartford Courant, .

FOUCAULT, M., 1977. Discipline and punish: the birth of the prison. Pantheon Books.

LYNCH, C., 2017. The rise of reading analytics and the emerging calculus of reader privacy in the digital world. First Monday, 22(4 (April 3rd)),.

MARTHEWS, A. and TUCKER, C., 2015. Government surveillance and internet search behavior.

MATHSON, S. and HANCKS, J., 2008. Privacy Please? A comparison between self-checkout and book checkout desk circulation rates for LGBT and other books. Journal of Access Services, 4(3-4), pp. 27-37.

PEN AMERICA, 2013. Chilling effects: NSA surveillance drives US writers to self-censor. New York: PEN American Center, .

PENNEY, J., 2016. Chilling effects: Online surveillance and wikipedia use.

Quad/Graphics v. S Adirondack 174 Misc.2d 291 (1997), 664 N.Y.S 2d 225. 1997.RICHARDS, N.M., 2008. Intellectual privacy. Tex.L.Rev., 87, pp. 387.

RICHARDS, N., 2015. Intellectual privacy: rethinking civil liberties in the digital age. Oxford University Press.

SCHAUER, F., 1978. Fear, risk and the first amendment: Unraveling the chilling effect. BUL rev., 58, pp. 685.

STOYCHEFF, E., 2016. Under Surveillance: Examining Facebook’s Spiral of Silence Effects in the Wake of NSA Internet Monitoring. Journalism & Mass Communication Quarterly, 93(2), pp. 296-311.