Online chilling effects in England and Wales

Open and free internet-based platforms are seen as an enabler of global free expression, releasing writers from commercial and space constraints. However, many are working without the assistance of an in-house lawyer, or other legal resources. This may lead to undue suppression of public interest material, with important implications for freedom of expression and the democratic function of media. Two online surveys among digital and online journalists in England and Wales in 2013 indicated that the majority of encounters with defamation and privacy law take place outside the courts, with few formally recorded legal actions. This was particularly evident in a sample of hyperlocal and local community publishers. In light of the results, this paper calls for a reappraisal of overly simplistic judicial and media applications of the chilling effect doctrine, in order to expose its subjectivities and complexities. Additionally, attention needs to be paid to global and cross-jurisdictional medialegal environments, in order to help develop better internet policy and legal frameworks for protecting legitimate expression.

many have nothing to do with journalism or blogging.Particular judicial and policymaking attention is paid to their effect on blogging and journalism, however, with view to protecting media freedom of expression and the reporting of public interest information.There are specific public interest defences used by journalists (notably, the Reynolds defence, recently replaced by Section 4 of the Defamation Act 2013) but they are not necessarily restricted to journalistic activity.In regards to internet-specific provisions, case law has developed over time to include internet publication, including blogging and tweeting, and the Defamation Act 2013 introduced, for the first time, a specific statutory provision for operators of websites.
While broadly operating within the same legal context as bloggers and other members of the public in England and Wales (save a few special protections), journalists may have access to more resources through an employer or be more likely to encounter legal action because of their perceived influence and the types of topics and people they cover.This is largely a supposition owing to the inadequate public records available to policymakers and researchers and the difficulties in collecting consistent data (see Townend, 2012 and2013).To get a clearer sense of bloggers' and journalists' actual experiences and perceptions of the chilling effect, I ran a number of surveys in 2013.The aim of these surveys was to gather data on internet-specific journalism and independent news blogging, especially those working without the support of mainstream media organisations.Previous studies have documented the experiences of print and television journalists (e.g., Barendt et al., 1997); this exercise attempted to look at the libel and privacy climate for those engaged in digital journalism at two levels: those working on a range of topics and at national level; and those concentrating on local and community news.
The online surveys among over 200 journalists and 'hyperlocal' and community news bloggers in England and Wales2explored their legal resources and support, the impact of libel and privacy on their work, direct legal experiences (such as receipt of a threatening letter), and their overall perception of the chilling effect over a five year period .The surveys were divided between a 'general' group of journalists and online writers reached through social media channels and industry websites, and a specific hyperlocal group (mainly running small community websites)3 targeted by email.While the sample size is relatively small and the general group self-selecting, general patterns in publishers' experiences of the law can be traced from the quantitative data, while the qualitative material offers a range of publishers' reflections.
The surveys expose a spectrum of interpretation; at one end respondents appear unaffected by libel because of their ignorance and lack of awareness of the potential risks; at the other there is even evidence of excessive self-censorship because of their legal knowledge and experience.
Respondents' view of the impact of these laws is dependent on a complex web of influences; as well as knowledge, their editorial decisions are dependent on the legal resources available to them and even their personality.Crucially, their experiences will vary depending on the type of topics they write about, the nature of the publication they are writing for and the medium.The manner in which they are affected varies radically too: they may have been sued, or received warning letters pre or post publication about a possible libel claim, or simply feel uncertain about the legal implications of publishing certain material.All of these factors contribute to their decision to publish or 'spike' material, or may even deter them from pursuing ideas in the first place.

INTERNET LIBERATION?
Open and free internet-based platforms are seen as a great enabler for global free expression (see, for example, Dutton et al., 2011;La Rue, 2011), releasing writers from traditional commercial pressures and the constraints of print space and broadcast airtime.However, many of these writers are working without the assistance of an in-house lawyer, or other legal resources provided by large media organisations (Townend, 2011).This can have both a liberating and censorial effect on speech: on the one hand, a small organisation or individual blogger may be perceived as less influential in opinion-forming and therefore less likely to get sued, leaving them freer to explore material; on the other, the absence of an advisor and limited financial means could cause them to unduly suppress ideas and material in anticipation of costly and time-consuming legal action.
Recognising the liberation from mainstream restrictions, one respondent suggests that nonlegally trained bloggers are able to fill "some of the gap" left by journalists afraid to publish material.However, in their view, the bloggers "may not consider the full consequences": Journalists are often "shown up" for "missing" public interest stories broken by bloggers, when it's likely the case that journalists may have certain information but a fear of publishing it in case of legal action.
Equally, a lack of training and support can cause unwarranted suppression of speech.Another respondent suggests that: I learnt the hard way that you cannot really write effectively as a citizen journalist without access to legal assistance -public and private organisations do not take criticism well.
The surveys show that those mainly writing for third party sites, rather than their own sites or blogs, are much more likely to have access to legal advice.In the general survey just under half of the journalists and online writers contributing to other publications have access to such services (45%).By contrast, under 10% have access to legal advice for their own publication or website.4A similar variance was seen in the hyperlocal sample, although most respondents run their own websites.
Media law insurance, offered by a small number of specialist firms, offers another form of financial protection in the event of legal actions being pursued, although the level of protection would vary depending on the terms of the cover.Of respondents who contribute to third party publications, a minority have, or are aware that they have, media law insurance for the main publication to which they contribute: 34% of the general sample and 28% of the hyperlocal group.5Only a very small number have such insurance for their own publication, either a personal blog or more substantial operation: 3% in both general and hyperlocal groups.6Cost seem to be a deterring factor for many respondents although some indicate they simply are not aware of the availability or price of such insurance or how it would help them.In fact, a few respondents think that having insurance would have the perverse effect of encouraging claims.
According to one, "If people know you are insured they will sue."However, as one respondent notes, even mainstream publications may not have libel insurance.
He or she had found out during a legal training session "that hardly any newspapers now carry legal insurance -cheaper to settle on a case-by-case basis than the actual premiums."This was certainly the case at the Guardian in 2011, when its editor Alan Rusbridger confirmed in a comment under the line that "we aren't insured for libel" (2011).Instead, media organisations may take out After-The-Event (ATE) insurance for a specific complaint, if they are able to secure it, or meet the costs of cases and settlements from internal budgets.Nonetheless, the absence of regular legal insurance can have a detrimental effect on publishing activity: investigative journalist Heather Brooke has described, for example, how one of the biggest obstacles faced by the Bureau of Investigative Journalism "was finding reasonable libel insurance".The founder of another online initiative, HelpMeInvestigate, faced a similar problem, when it took out libel insurance to satisfy its original funders but once operating independently found that the costs were prohibitive.7According to Brooke this "legal nightmare halts small or online cooperative journalism sites in their tracks."(2012)

TO PUBLISH OR SPIKE?
The survey responses support anecdotal evidence that many independent bloggers and online writers publish material without the resources of mainstream organisations.The next question, then, is how this affects their editorial decisions.Are they, as has already been suggested, liberated by their limited infrastructure or does it cause them to withhold information that they wish to publish?Respondents were asked about the way in which libel and privacy law impacted editorial decisions to change or abandon stories that they considered to be in the public interest.
It is the latter decision that is often represented as one of the most potent chilling effects of defamation (see e.g., Glanville and Heawood, 2009).
The general sample indicates that they are more likely than the hyperlocals to substantially change stories they consider to be in the public interest because of libel law, with around six in ten saying they change stories "some of the time" (62%), while around five in ten of the hyperlocals say they do this "none of the time" (53%).8In terms of abandoning public interest stories, over half of the general sample said they do this "some of the time" (57%).A slightly bigger proportion of the hyperlocal sample say they drop such stories because of libel law "none of the time" (61%).9 The same questions were asked of privacy law.Privacy law seems to have a lesser impact, with only one third of the general sample (35%) saying they substantially change public interest stories because of privacy law and over two thirds of the hyperlocal sample saying they never do (70%).10 Broadly speaking, hyperlocals are more likely to say that they change or abandon stories because of libel and privacy "none of the time" and the general sample "some of the time".The most marked contrast, when splitting the samples into sub-categories, was between (a) journalists from the general sample mainly publishing on other people's sites and (b) non-journalist hyperlocals with their own publication.For the first group (a), six in ten say they abandon stories "some of the time", whereas seven in ten in the latter group (b) say they abandon stories "none of the time".11This indicates a greater tendency to abandon stories for reasons of libel among journalists working for an employer or third party sites, than for hyperlocal nonjournalists working for themselves.It appeared, somewhat perversely, that the greater the level of legal resources and training, the more likely they appear to be affected by libel law.

BENEATH THE SURFACE
The survey then turned to actual experiences with the law, asking respondents about the number of threats they had received, and how many of these led to formal legal actions.Are the lower resourced bloggers more or less likely to experience litigation, or threats of legal action?Among respondents to the general sample, 46% say they received threats of libel action over a five year period , in comparison to 35% of the hyperlocal sample.12Respondents who answered positively were asked how many such threats they had received in the last five years.
The mean was five such threats among the general group and six among the hyperlocal group although the ranges were large, with one respondent from each sample reporting over 50 such instances alone -a clear anomaly.Similar questions were asked about the threat of privacy action.Far fewer had experienced threats of breach of privacy and confidence than libel.Of the general group only 15% had received such threats, and 11% of the hyperlocal group.13 The surveys suggest that the majority of encounters with defamation and privacy law take place outside the courts, with few formally recorded legal actions brought against publishers, supporting Barendt et al.'s suggestion in the mid 1990s that actions in court only represent the very tip of the iceberg (1997, p 41).In fact, a minority of the hyperlocal group experienced threats of legal action and none were reported to have reached court.Only 12% of the general group who had experienced threats report that formal legal action was pursued against them in the stipulated five-year period.Of the seven cases reported by four respondents, three settled with payment of damages and/or costs, one settled without payment being made, one was resolved with an offer of amends, and two were decided at trial.Although there were far fewer privacy than libel claims, a similar pattern emerged where very few, if any, ended up as a formal court claim: in the general group, only one respondent reported that a privacy threat had led to a claim issued in court.In the hyperlocal group, no libel or privacy threats led to a claim being issued in court.
The questionnaire did not ask the general sample about the outcomes of threats that did not reach court, but it was pointed out in an email from one respondent that this missed out threats that were settled after content was removed or changed and/or a payment made, without any formal legal claim ever being issued.In response, a question was added to track outcomes in the hyperlocal sample.This revealed precisely what the correspondent suggested: that unofficial claims were being resolved in ways that did not involve court, with 76% of libel threats not pursued further by the complainant.Interestingly, no outcomes were reported to have involved payments.Of 148 outcomes cited by 24 hyperlocal respondents, 32 were settled out of court with an apology and/or correction and/or removal of content (and no payment); 113 were not pursued further by the complainant; two are still on-going and one had another outcome.
Of course, the numbers are small here, and the respondents were citing incidents from memory, but it serves as a useful indication of what happens to legal threats that are never formally recorded in the courts system.During the period of research other journalists have described similar incidents14, where a complaint has been resolved by a payment or removal of the content "in full and final settlement" of the matter.
The data indicates that the number of journalists and bloggers changing or abandoning material is greater than the number actually receiving threats of legal action, with a small minority experiencing a formal claim issued in court.This would appear to mirror what has been said about regional newspapers: they are concerned about libel and privacy but are very rarely engaged in defending formal complaints (Rasaiah, 2013).A lack of incidents in court does not mean that digital journalists and writers do not think about, or react to the pressures of defamation and privacy law, either as the result of direct threats or anticipated threats.

PERCEPTIONS OF THE CHILL
What is more difficult to establish is whether libel and privacy law cause undue suppression or deterrence of public interest material; this is what Schauer has identified as an invidious chilling effect, where libel and associated costs inhibit protected expression beyond the intention of the law (1978, p 690).While the questionnaire attempted to be as specific as possible and asked respondents to consider whether 'public interest' material was substantially changed or abandoned for reasons of libel and privacy, assessments were inevitably informed by respondents' subjective positions.It would be fruitless to attempt an objective assessment of chilling effects when their component parts -such an individual's perception of what they should be legitimately allowed to publish -are moving and varied.
It is possible, however, to collect and categorise the ways in which publishers' perceive a chilling effect and the detrimental impact of libel and defamation law on publishing activity.The survey asked whether respondents recognise a chilling effect on their work, and if so, what caused it and how it manifests itself, followed by a question on whether libel and privacy law prevents the publication of material in the public interest.Some simply do not recognise a chilling effect, perhaps because of the type of content they produce, with one respondent suggesting his or her perception might be due to a lack of knowledge.For another respondent, it is the reverse: their specialist knowledge of defamation and privacy prevents them from "even thinking" of the type of stories that would be subject to a chilling effect; it is not clear if they consider this a good or a bad thing.
Others do not recognise the negative effect of libel and privacy on their work, but reflect on the various forms of a perceived chilling effect that might be experienced by others.For example, one respondent suggests that journalists in senior roles might feel the chill more keenly than those lower down the food chain: I can't think of an instance where that [the chilling effect] has been the case, but there is a strong possibility that knowledge of how things have changed alters depending on what level you are at in a newsroom.E.g., a news editor might respond that they are "spiking" more stories that are being pitched to them, because they know publication would be ultimately unlikely.
A handful of respondents point to the other factors, besides libel and privacy, that lead to the abandonment or redaction of stories, for example: the "negotiation between sources and journalists"; an editor's aversion to an overly polemical tone; and external and internal commercial pressures relating to advertising and business relationships.For many, the detrimental effect of defamation and privacy law is felt strongly, especially in relation to the unequal financial resources between claimant and defendant: Financial clout is a bullying tactic used in libel cases by the wealthy.Cases which should never even be raised are "won" (normally settled out of court) not because the claim is just but because the threat of ruinous legal costs if defended/unsuccessfully is too dreadful to contemplate.It's like playing poker with somebody who can afford to raise you until you are in tears at the table and have to fold.
Another common concern is the difficulty in acquiring the necessary evidence to defend a potential libel complaint.For example, There is a chilling effect in that we come into possession of information from more than one source about an issue in the public interest that will still not have enough documentation to be defended in a libel action.The threat of libel is a deterrent in itself because of the time and expenses which we as a small organisation cannot afford.My main response is to ensure that all information is well attested to and documented before publication, though due to time and expense pressures this means that some stories cannot be pursued to their full extent.
Answers from the general sample tend to be lengthier than those from the hyperlocals, indicating that respondents in the general group spend more time dealing with, or considering the issues at play.But in both samples, responses are varied, with the chill perceived at different points of the process, subject to different influences, and of varying levels of concern.
There is, however, one unequivocal conclusion that could be drawn from the results in this section: there is no one 'chilling effect'.Despite its generalised use in relation to libel in media and judicial discourse, it clearly means different things to different people.While the chilling effect is very real to some writers, they interpret it in different ways, offering definitions based on variable components, such as access to resources, legal knowledge and personal experience.Furthermore, the 'chill' is perceived at a variety of stages of the editorial process, directly and indirectly, or 'structurally' (see Barendt et al., 1997).This suggests the climate is not universally chilly for publishers in England and Wales; it can be more confidently described as hazy, with some people feeling the cold more than others.
The distinct ambiguities of the chilling effect call for further study, rather than less.While its complexity has already been acknowledged in scholarly work (e.g., Schauer, 1978 andKendrick, 2013) there is scope for further examination of the subjectivities and complexities at play in freedom of expression negotiations and disputes, by gathering new empirical material and building on past research conducted, for example, by the Iowa Libel Project (Bezanson et al., 1985), Newcity (1991), Dent and Kenyon (2004), Cheer (2008) and Baker (2008).Such a theoretical exercise would have practical purpose as well.A more nuanced definition and recognition of the moving component parts of the chilling effect would help inform on-going policy making around media dispute resolution, libel and privacy, as well as broader issues related to online freedom of expression.It would also help establish that policy initiatives designed to eliminate the 'chill' would be futile given its subjective form; instead a more realistic policy aim is to help reduce the illegitimate deterrence of freedom of expression with more effective systems for resolving disputes and dissuading claimants from bringing or threatening unfounded claims.Furthermore the impact of the current law needs to be carefully monitored to assess whether the existing law is fair and proportionate for all parties and, if not, inform the development of future legal reform.

USING THE FINDINGS TO INFORM POLICY DEVELOPMENT
How might the findings of the online surveys be used to inform policy making in England and Wales, and other comparable jurisdictions?

LIBEL AND PRIVACY DISPUTE RESOLUTION METHODS
It is clear that the perceived chill occurs outside the courts and is largely undocumented.While claimants and lawyers' understanding of the substantive law affects the nature of complaints brought, attention also needs to be on the procedure of dealing with reputational and privacy disputes outside the courts system.However, recent policy making addressing the chilling effect of libel law has concentrated on the substantive law through the creation of a new statute, the Defamation Act 2013.While procedural aspects are also under review as part of the government's attempt to reform costs and the on-going development of a new system of press regulation, these endeavours largely concentrate on large media publications, defined in statute and by the UK Department of Culture, Media and Sport as "relevant publishers".
The surveys presented in this paper examined the experiences of bloggers and small-scale publishers whose freedom of expression rights also deserve robust protection, especially as mainstream media organisations cut back on important activities such as court reporting, specialist local coverage and investigative journalism (see, respectively, Rozenberg, 2009;Ponsford, 2012;House of Lords, 2012).A key policy question, then, should be how to develop dispute resolution methods for bloggers and small publishers that would help avoid the risk of expensive litigation and deter illegitimate claims or unnecessary and hasty settlements.Various forms of mediation, arbitration and Early Neutral Evaluation (ENE) are considered in the Alternative Libel Project report (2012, PDF) but do not appear to have been fully explored by policymakers.An additional policy option that might be explored in the UK is specific legislation to deter 'strategic lawsuits against public participation' (SLAPPs), although designing appropriate legislation would present a number of significant challenges, not least because of the subjectivities at play when determining the legitimacy of actions (see Scott, 2011).

PUBLIC LEGAL EDUCATION INITIATIVES
The surveys revealed that having a low level of resources does not necessarily lead to a heightened perception of the chill and over-restriction of public interest material, nor increased numbers of threats and libel or privacy suits.One reason for this could be that a significant proportion of bloggers and small-scale publishers (especially in local contexts) are less likely to write about legally sensitive topics, or people with the resources or inclination to sue.Separate research in 2013-14 on over 180 hyperlocal media outlets indicates that many local bloggers are doing important and legally risky reporting on topics such as local government corruption, but that this is a minority exercise, with only four in ten indicating that they have started a campaign or investigation in the last two years (Barnett and Townend, forthcoming).
Additionally, bloggers might avoid libel threats because they are perceived to have less public influence than mainstream media, but this is not guaranteed.However, this does not mean small publishers receive any formal protection from libel and privacy claims, and the effect of even one such threat or action could jeopardise the future of their operation.This, coupled with the fact that some answers indicated ignorance or misconceptions about the reach of the law, necessitates the creation of better support and training systems.Knowledge and advice can help prevent an inadvertent libel, but also enable an uncertain writer to act within the full limits of the law and avoid excessive self-censorship.To help protect bloggers and small publishers, especially in local and community settings, policymakers would do well to consider how libel and privacy training might be developed in different educational settings, through schools, universities and non-profit organisations.There could, for example, be opportunity for the BBC to provide legal education and general information to community publishers, as part of its development of online and creative partnerships (see Hall, 2013).The government and other relevant agencies could also consider making funding available for public legal education projects, specifically teaching members of the public about libel and privacy law.

OPPORTUNITIES FOR COMPARATIVE RESEARCH
Finally, while the surveys concentrated on activity within one particular jurisdiction, England and Wales, further attention needs to be paid to inter-jurisdictional and global media-legal environments.This research concentrates on journalists and bloggers publishing to a substantial audience based in England and Wales, but they are not immune from actions and threats brought in other jurisdictions, in and outside of Europe.Similarly, journalists and bloggers based in other jurisdictions may face threats from England and Wales, and claims may be brought if a plausible and significant connection to the jurisdiction is established.Future research would do well to consider the challenges of the globalised media-legal space, in order to help develop better internet policy and legal frameworks for protecting legitimate expression and the public's right to impart and receive information.

CONCLUSION
Drawing upon empirical research findings, this article has considered the development of internet policy and legal frameworks to protect legitimate expression and the public's right to impart and receive information.Too often, policymakers focus on the tip of the iceberg, through anecdotal evidence and inadequate formal records (Townend 2013), which misses what is actually happening beneath the water.This means that policy attention is often drawn to unusual predicaments (e.g., court litigation) and fails to address the problems associated with everyday activity (such as receiving an empty threat, or attempting to reach an out-of-court settlement with limited legal resources).While the quantitative findings of these modestly sized surveys must be treated cautiously, wider research aided by industry and judicial resources and backing, would be in a position to test their conclusions.There is scope for detailed comparative research in different jurisdictional contexts which help inform legal and regulatory policy making in globalised and digital spaces, as well as scholarly and judicial examination of rights related to reputation, privacy and freedom of expression.