By way of preface this is a paper that followed a presentation at the Criminal Bar Conference in Auckland in 2018. It was submitted for publication the the NZ Criminal Law Review – the organ of the Criminal Bar Association – but continued delays have meant that the paper has not seen the light of day.
I have never had to publish or perish for Performance Based Research Funding and the only benefit that I can see with academic publishing is that a piece gets peer reviewed. One of the things that the Internet allows is for an automatic peer review process to take place once a blog post is put up. The upshot of all this is that I think I shall use this blog for my academic as well as non-academic pieces. As far as quotability or peer reviewing is concerned – as Caesar said – iacta alia est.
A copy of this paper has also been posted on Scribd – https://www.scribd.com/document/419303736/Diluting-Prejudice
This paper is about the steps that may be taken to dilute but not totally eliminate the prejudicial effect of on-line material that relates to a defendant about to stand trial. It is posited upon the fact that jurors, despite strong judicial directions, will go on-line to seek out information relevant to the case that they are trying. The solution lies in the use of what could be described as “take-down” orders whereby material is removed from websites and de-indexed from search engines during the course of the trial to eliminate or dilute any prejudice that may otherwise arise. The remedy of a “take-down” order restores the qualities of practical and partial obscurity of prejudicial information that was a characteristic of the pre-Digital Paradigm. The paper also discusses the issue of juror contempt in light of proposed changes to the law in the Administration of Justice (Reform of Contempt) Bill.
In my article “The Googling Juror: The Fate of the Jury Trial in the Digital Paradigm” I considered the challenges posed by the Internet to the criminal jury trial. The Internet has dramatically changed the way in which people obtain, use, share and relate to information. As a result, it has become more difficult to shield jurors from extraneous information during trial and, as I observed, it is easier for jurors to undertake their own research or share information about a trial beyond the courtroom. Visiting a scene can be done virtually by using Google Earth or Google Street View. Such information is increasingly readily available on the Internet via a computer or a smartphone.
I referred to a suggestion that to address the problem of juror attempts to access online information relevant to the trial, lawyers could conduct their own Internet research in advance to identify what information about the case is available, analyse it and deal with it during trial. I observed that Courts in dealing with applications for severance and change of venue evaluated pre-trial publicity and these practices could provide a possible framework for evaluating new online sources of information that courts and jurors might access before or during trial.
This article considers another remedy that may be available to deal with highly prejudicial information that may be on-line relating to a trial or to an accused. A “take-down” order may be made, directed at an online content host requiring the removal of prejudicial content during the course of the trial. Associated with such an order may be a requirement for the de-indexing of the material from a search engine such as Google, again during the course of the trial. It is acknowledged that such orders will not provide a complete answer to the problem, nor would an order for suppression pursuant to the provisions of the Criminal Procedure Act 2011. However they will serve to dilute the possible prejudicial effect accompanying pre-trial publicity.
The article commences with some observations about aspects of pre-trial publicity and the ability to recall such information in the pre-Digital Paradigm. It will be suggested that in this pre-Digital informational environment, given the various obstructions to the speedy recovery of archived information, compliance with judicial directions to jurors to refrain from carrying out their own investigations was likely to be greater.
Consideration will move to how the Internet challenges those previous assumptions about information, why it is that jurors are able to ignore judicial directions and why they might be likely to do so. This discussion will reflect on recent examples from cases where such directions have been ignored.
The discussion will then turn to recent cases involving take down orders. A number of cases will be considered with a focus upon four of them and the themes and assumptions arising from them. An attempt shall be made to ascertain why in these cases it is assumed that judicial directions will reduce the likelihood of jurors carrying out their own Internet-based investigations. It will be argued that this confidence is misplaced and that pro-active judicial information management is necessary to protect the right to a fair trial in the Digital Paradigm.
The importance of the evaluation of the content the subject of a take-down order will be considered with the tension that exists between the freedom of expression and the right to a fair trial as a backdrop.
Some of the proposals of the Law Commission and the use of court orders for a take-down of the material as a prophylactic to contempt will be contemplated .
There is now no doubt that the Court has the power to make take down orders. Practical considerations will be addressed including the procedure that should be adopted, those who should be served and heard on a take-down application and the type of information that should be put before the Court. The attitude of Google LLC to compliance with domestic Court orders has been the subject of some media publicity; a possible solution to this apparent difficulty will be suggested.
The article will conclude with some thoughts about some possible futures for take-down orders as we move further into the Digital Paradigm.
II. The Nature of Information and the “Fade Factor”
As the passage of time dulls memory, the impact and freshness of a news report is lost. This has been judicially recognised in cases where there has been a high level of publicity. Examples may be found in the cases of R v Bailey and Iti v R which both involved suspected terrorist activity in the Ureweras which attracted a high level of media interest. The effluxion of time meant that jurors might recall some of the publicity but not to the extent that a fair trial would be prevented. In R v Reddy the Court, in noting that retrials may be held in the same location as the original trial, referred to the “fade factor” 
“because any attendant negative or prejudicial publicity is presumed to have faded from potential jurors minds. The collective memory of the public is subject to a “fade factor””
Before the advent of the Internet in what may be termed the analogue or kinetic paradigm, the distribution and dissemination of information about a police investigation or a particular crime was almost exclusively through the news media – newspapers, radio or television. Despite the fact that television is seen as an influential medium, pre-Internet, it relied on what could be termed an “appointment viewing” model. Like radio, the audience had to be present in front of a receiver to listen to or view the broadcast. Once the broadcast had taken place, unless it was repeated, that was the only chance the listener had to hear or see the content.
Newspapers and magazines were slightly more enduring and could be read and re-read at leisure. However, the long term retention of copies of newspapers or the articles published was left more to the “morgues” of the newspapers themselves or the archives of a local library.
Another aspect of information dissemination was the way in which mainstream media was organised as a business. The method of dissemination for newspapers, radio and TV was (and generally still is) from a centrally located conglomerate –a radio or television studio or a newspaper publishing facility – to a widely distributed audience. This allowed for the control of the content and flow of information from the media organisation to the public audience.
The presence of these pre-Internet factors presented obstacles to the retrieval of historic pre-trial publicity about a case that enhanced the “fade factor”. These obstacles can be described as practical and partial obscurity.
- Practical Obscurity
Practical obscurity refers to the quality of availability of information which may be of a private or public nature. Such information may be indexed, in a central location or locations such as public libraries or newspaper rooms, in hard copy format, and is frequently location-dependent in that the information will refer only to the particular area served by that location. Interaction is required with librarians, officials or bureaucrats to locate it and to an extent prior knowledge of the particular file or repository within which the information source lies is required.
Practical obscurity means that information is not indexed on key words or key concepts, but generally on the basis of individual files or in relation to a named individual or named location. Thus, it is again necessary to have some prior knowledge of information to enable a search for the appropriate file to be made. These are obstacles to the ready access to information. 
One practical example is the operation of hard copy newspaper archives. The ability to access a back copy of a newspaper poses a number of obstacles: travel to the repository by public or private transport, locating the volume within which the newspaper may be located and then laboriously reading through each copy of the newspaper until the article is located. The process is further complicated by the fact that hard copy of old newspapers is no longer available and the copies of newspapers are on microfilm.
- Partial Obscurity
Partial obscurity addresses information of a private nature which may earlier have been in the public arena, in a newspaper, television or radio broadcast, or some other form of mass media communication. Later than information can only be recalled in part as the result of the inability of memory to retain all the detail. Thus, a broad sketch of the information renders the details obscure, only leaving the major heads of the information available in memory, hence the term partial obscurity. This underpins the concept of the “fade factor” referred to above.
To recover particulars of the information will require resort to film, video, radio or newspaper archives, thus bringing into play the concepts of practical obscurity. Partial obscurity may enable information which is subject to practical obscurity to be obtained more readily because some of the informational references enabling the location of the practically obscure information can be provided.
These two factors exemplify the logistical difficulties confronting a would-be “investigative juror.” The “fade factor” worked to dilute and reduce the impact of any potentially prejudicial publicity so that whatever memory a juror might have gathered from pre-trial publicity about the circumstances of a case would have significantly reduced. Thus the clear recitation of events in the formal court-room setting would be far more likely to supplant any memory based or vague recollection of events. A direction to focus upon the evidence was far more likely than not to receive compliance simply because of the difficulties a juror might encounter in trying to locate earlier reports of a case.
This is not to say that the news media never overstepped the mark and published material that was prejudicial. It is for this reason that a remedy lay for publication contempt where there was a real risk, as distinct from a remote possibility, that a publication interfered with the right to a fair trial. The strength of the test demonstrates that the nature of the publication might be likely to override the “fade factor” or the effects of partial and practical obscurity.
C. The Internet as a Problem
The Internet challenges these concepts and indeed the “fade factor”. One writer has characterised the Internet as “digital memory”. To understand the nature of the challenge, it is necessary to briefly sketch the topic of information qualities. These qualities have been developed to distinguish digital information from that of the pre-digital era. Information qualities sit below the content layer and involve a consideration of the medium of communication. In this way, McLuhan’s aphorism “The Medium is the Message” comes into sharp focus.
It would be wrong to say that the qualities of digital information are completely novel. Some are present in the pre-Digital Paradigm but as new technologies have become available these qualities have been enhanced. For example the quality of dissemination that Elizabeth Eisenstein argued was one of those that characterised and differentiated print technology from those of the scribal culture is present in the Digital Paradigm but to an extent unimagined in the print paradigm, limited as it was by the physical nature of copies.
Along with the quality of exponential dissemination, two other qualities of digital information technologies – information persistence and information searchability\retrievability – especially highlight the paradigmatic difference that online information presents to the earlier Kinetic Paradigm.
Information persistence recognises that once information reaches the Internet it is very difficult to remove. It spreads through the network of computers that comprise the Internet and may be retained by any one of them. It has been described as the phenomenon of “the document that does not die”. Although information may be difficult to locate on the Internet, information persistence means that it will be available somewhere, if only in an archive. The fact that information is persistent means that it can be located by the digital equivalent of an archaeological dig – except that the trowel and spade are replaced by a search engine, which brings us to the searchability\retrievability quality.
Searchability\retrievability falls within the classification of user associated qualities, although there is a technical aspect to it as well. The technical aspect lies within the makeup of digital information. That information is in digital format which means that it can be searched. This is in startling contrast to information in documentary form which must be read – what is referred to as manual review – to retrieve relevant information.
Electronic discovery demonstrates the way in which the machine itself provides an answer to a machine-based problem. In litigation, huge volumes of digital information require analysis to determine the files or materials that are relevant to the case in question. To print out what often amounts to tens of thousands of pages, which then would have to bee manually reviewed, is seen as disproportionate in terms of time and cost. Software tools are thus deployed in e-discovery exercises, built upon the premise of quality of searchability of digital information.
The most ubiquitous search tool on the Internet is Google, but the same model underlies all search engines.
Search engines consist of 3 main parts. Search engine “spiders” follow links on the web to request pages that are either not yet indexed or have been updated since they were last indexed. These pages are “crawled” and are added to the search engine index (also known as the catalog). When the user searches using a major search engine, what in fact is searched is a slightly outdated index of content which roughly represents the available indexed content of the web. The third part of a search engine is the search interface and relevancy software. The search engine adjusts the search query for spelling variations, checks to see if the query is relevant to other vertical search databases and gathers a list of relevant pages, ranked according to the parameters in the page ranking software.
Search engines are essential for the proper functioning of the Internet. Without them, the information that is located in servers on the network would be largely inaccessible unless the user was aware of the location of that information.
The basic search using a search engine is one way of obtaining information required. Electronic material can be cross-referenced and indexed according to a number of criteria and may be selected on the basis of content as well as other identifying information. Using a full-text search, it may be possible to pinpoint information that may not be returned using standard keyword or metadata based searching. Indeed, if court decisions and records are open to web crawlers or web mining, past case information may be retrieved.
Thus it may be seen that the concepts of partial and practical obscurity are overturned by the qualities of online digital information. Indeed, prejudicial information about a defendant, his previous criminal activities, associations and possibly even his convictions can be made available to an enquirer with ease.
In addition, the Internet reverses the “flow of information”. In the model of partial\practical obscurity the enquirer was required to go to the information; in the Digital Paradigm the information flows to the enquirer. Furthermore it has the same immediacy of the original publication had and is presented in “news” format. This “replication and recovery” of earlier news information has, depending upon the content, the potential to be highly prejudicial to a defendant’s “fair trial” rights.
The concerns that were expressed particularly by Wylie J about “historic” information and the lengths that a juror might have to go to locate prejudicial information overlooks a number of matters.  First, as has been observed, the Internet allows for the preservation of information so that when it is read it is as fresh as the day it was first published and its impact is maintained. Secondly, search engines enable the recovery of this information. As has been suggested, the Internet challenges the concepts of partial and practical obscurity. A “googling juror” need only search on identification particulars that are raised in the course of the trial to locate information. With respect, Wylie J probably underestimates the sophistication of search engines, their ability to retrieve information and the skill of an ever-widening community in being able to locate Internet based information. The emphasis on historic material is misplaced. If it is on the Internet, it is retrievable with the assistance of a search engine.
III. Do Jurors Go Online?
A. Internet Accessibility
Given the accessibility of information via Internet platforms, the likelihood of jurors conducting online researches is increased. To understand the nature of the problem it is necessary to appreciate the accessibility that New Zealanders have to the Internet.
The Institute of Culture, Discourse and Communication at Auckland University of Technology has conducted a number of surveys of Internet use in New Zealand under the name of the World Internet Project New Zealand (WIPNZ).
The fourth survey of WIPNZ was conducted between late July and early September 2013. In the Executive Summary the following observation is made about Internet usage on the part of the 2006 New Zealanders surveyed.
“For a large number of people the internet is used daily. Four out of five spend an hour or more online at home every day. Almost everyone under 40 is online, so that only 1% of our under-40 sample are non-users. Accessing the internet ‘on the go’ is prevalent. Seven out of ten users access the internet from a hand-held mobile device such as a smartphone or an iPad. Almost half of the internet users surveyed (48%) said that they had accessed the internet through a tablet, while an even higher proportion (68%) connected through their mobile phone in the past year.”
The findings of the 2013 report indicate that Internet usage by a large sector of the New Zealand community is routine.
The 2015 WIPNZ report continued to monitor Internet usage patterns. Of those surveyed, only 8% did not use the Internet and were divided into ex-users (3%) and never-users (5%). This means that of those surveyed in 2015 92% were Internet users. Only 11% of the sample were described as low level users.
The most recent survey indicates that 93.8% of the population have Internet connections but only 80% have a home connection. There are some 3.85 million mobile Internet connections – a figure which has stayed relatively steady with fluctuation over the last five years between 82% in 2013 and 79% in 2015 and 2017. Improved infrastructure and the introduction of ultrafast broadband has resulted in increased upload and download speeds and the uptake of fibre is fuelling large rises in data being used across New Zealand networks.
The importance of this statistical information is that it demonstrates how Internet access and use has become part of the normal life of New Zealanders. It can be deduced from this that most, if not all, members of a jury pool will be Internet users, accustomed to Internet use and comfortable with obtaining information from the Internet.
B. Overseas Juror Misconduct Studies
Having established that Internet access and use is a routine part of the lives of a very large number of New Zealanders, the discussion turns to a consideration of the use of the Internet by jurors. It is submitted that this is wider than may be initially thought, although no empirical research has as yet been conducted in New Zealand.
However, juror use of the Internet has been studied in England by Professor Cheryl Thomas of University College London in a study undertaken in 2010 for the Ministry of Justice entitled “Are Juries Fair?” Professor Thomas’ study was conducted in three different locations (Nottingham, Winchester and London) and included 62 cases and 668 jurors. The sample included both long, high profile cases and standard cases lasting less than two weeks with little media coverage. Her findings revealed that those jurors who did seek out information did so using the Internet. Interestingly enough, more jurors said that they saw information on the Internet than those who admitted looking for Internet based information. The jurors admitted that they were doing something they had been told not to do, which may explain why more jurors said that they saw reports than those who admitted looking on the Internet. There was a higher incidence of Internet enquiry in high profile cases.
81% of those who in these cases sought Internet-based information were over 30. Of all the sample who said they sought Internet-based information, 68% were over 30. Professor Thomas’ study demonstrated that the problem of “The Googling Juror” is not limited to younger jurors. 67% of the jurors in Professor Thomas’ study were between the ages of 30 and 59. 17% were within the 18 – 29 year old age bracket in the Nottingham Crown Court study. The figures were 59% and 18% for the Winchester Crown Court. Thus the majority of jurors were over the age of 30.
In a subsequent article Professor Thomas suggested that her research revealed that a small minority of jurors did not follow the rules relating to juror Internet use.  She considered that conditions could amount to a “perfect storm” of improper juror conduct where jurors did not understand that they should not look for information (via the internet or elsewhere) about their case during the trial; that when jurors find such information they share it with other members of the jury; and where, even if other jurors know this behaviour is wrong, they are unwilling or do not know what to do to ensure that any verdict they return is fair.
In observing that it was impossible to monitor all aspects of Internet use during a trial, Professor Thomas was accepting that there would be some incidents of juror misconduct involving Internet use. Her position, in light of reviews that were proposed at the time of her article to English contempt of court laws, was that jurors need to understand what improper jury conduct is. Secondly, jurors need to clearly understand that if a fellow juror uses the internet improperly or if any improper conduct occurs it must be reported to the court. Thirdly, jurors must understand exactly how and when to report improper jury conduct and be provided with guidance that enables them to do so with ease.
Finally, Professor Thomas called for greater empirical evidence surrounding juror behaviour and what the best tools may be to assist them in performing their role. She urged that any reform of the law surrounding juror contempt should be based not on anecdotal evidence or high profile cases, but upon the fruits of such research.
- The United States
In the United States of America, one of the first surveys of jury behaviour was carried out by Professor Thaddeus Hoffmeister. This survey was somewhat wider than that of Professor Thomas in that it was sent to federal judges, prosecutors, and public defenders to learn how they viewed the impact of the Digital Age on jurors. The questions focussed primarily upon juror research and sought to assess the extent of the negative impact (if any) of the Digital Paradigm on jury service. 10% of respondents reported personal knowledge of juror research although it was acknowledged that this sort of behaviour was difficult to detect and probably under-represented the actual number of jurors who resorted to Internet-based research.
In an article in which she explores emerging technologies and its effect upon electronic juror misconduct, Judge Antoinette Plogstedt gathered together a large selection of instances of juror misconduct which had come to the attention of the Court.
The problem of what is referred to as “independent juror research”, or as we might call it “juror misconduct”, is the focus of an article by Robbie Manhas in which he argues that more liberalised procedural and evidentiary rules should be adopted to allow jurors to take a more active role in proceedings.
Assessing the frequency of juror misconduct relies, as was the case in Professor Thomas’ study, upon self-reporting.
“In a preliminary study of the frequency of juror and jury use of new media, the National Center for State Courts (NCSC) found that sizeable portions of trial jurors reported interest in using new media to conduct research on case-related topics and to communicate with friends and family about their jury service experience. Although the vast majority of jurors in that study had daily, if not immediate, access to new media, none of them admitted to acting on that desire. That study involved a very small sample of trials; however, it is clear from court opinions and news stories discussing the problem of the “Googling juror” that the risk is not purely hypothetical. In a review of court opinions published between 1998 and 2010, Thompson-Reuters reported that at least 90 verdicts were challenged based on claims of Internet-related juror misconduct. One-half of those challenges occurred between 2008 and 2010.”
The National Center for State Courts developed a study in 2012 to explore the impact of new media on juries, and to establish the necessary survey and other methodologies needed to do so. A pilot jury study was undertaken in 15 civil and criminal trials. Few jurors reported committing misconduct of any kind. However, a substantial portion either could not recall judicial prohibition on new media use or believed such searches were permissible. A sizeable proportion of jurors indicated a desire to use the Internet to obtain information relevant to the trial and a significant proportion indicated they would be unable to refrain from Internet use during the trial. Although the researchers were optimistic that the frequency of juror misconduct might be less than originally thought, the findings were less optimistic about the future. The vast majority of jurors had both technological access and the practical experience to use communications devices effortlessly and view these technologies as commonplace tools. The researchers observed that as younger cohorts join the jury pool access to the Internet and reliance on technologies for information will increase. They concluded
“A key factor will be the degree to which jurors continue to believe that the testimony of witnesses, especially expert witnesses hired by the parties, is more compelling evidence than what they can uncover on their own through information available to them via the Internet. Such conclusions will not be welcome news to those who wish to rely upon a more vigorous use of standard admonishments or on depriving jurors of access to the new media to keep the traditional, “unwired” jury.”
In an earlier article I made the distinction between “information in” where jurors seek information about a trial or related matters, and “information out” where jurors communicate information about the trial or their experiences on social media. Although the latter form of communication may prove ex post facto evidence or prejudice, the real concern must be the search for information relevant to the trial.
Two surveys, although recognising that Internet-based juror misconduct may occur, conclude that jury instructions are the most effective tool to mitigate the risk of such misconduct. The concerns surrounding those surveys were directed to wide issues of juror social media engagement during the course of a trial including communications during deliberations, communications with Facebook friends and jurors who blogged about their experiences. Although the surveys included incidents of juror research about a particular case, they covered a greater scope of misconduct.
The evidence is clear that there is a very real potential for jurors to go online and the possibility of a juror coming across prejudicial information (if it is available) is high. The number of incidents reported, especially from the United States, must give some cause for concern. Apart from the pervasiveness of and familiarity with the technology, it is suggested that there are deeper seated answers to the question why it is that jurors go online despite directions to the contrary.
IV. Why Do Jurors Go Online?
Why is it that some jurors are prepared to ignore or flout judicial directions and carry out their own researches online? This is a question the answer to which is not clear and the detailed study by Professor Thomas does not conclusively answer. In pre-Internet days, juror researches involved physical engagement, either seeking information from a library or newspaper room or visiting a scene. These examples of practical obscurity are challenged by the Internet.
A. Reversed Information Flows
The nature of information flows may serve to assist in clarifying the point. One of the factors that led to the practical obscurity of information was that the enquirer had to go and seek the information out. Thus the information flow was enquirer to information.
The availability of information online reverses that flow. Now the information moves towards the enquirer. There is little effort required, beyond carrying out a Google search, to seek out the information.
Coupled with this factor is that the enquirer is able to access the information from the privacy of his or her own home. This then leads to the way in which information can be sought surreptitiously. This quality of detachment may further explain why jurors are prepared to go online to seek out information.
Why is this significant? Juror enquiry is a recognised and recurring phenomenon. The consequences of such enquiry may be prejudicial to a fair trial and will result in mistrials and added delays in obtaining a fair outcome. Whilst an “after the event” solution such as a finding of contempt may contain certain deterrent qualities, a recognition of the phenomenon and an understanding of why it occurs will assist in considering and crafting possible remedies. The solution of a take-down order suggested in this paper, whilst not a cast iron one, will at least place obstacles in the path of juror enquiry.
Some of the drivers for juror enquiry that are inherent in digital information systems are now discussed.
B. Because They Can
One of the disturbing realities of the Digital Paradigm is the ease with which information may be obtained. The time is well past where a requirement for Internet access is dependent upon a laptop or desktop computer. Access may be effected by means of a smart phone or a portable tablet using a wireless connection.
The instant availability of the resources of the Internet coupled with powerful search engines such as Google, Bing or Yahoo puts information in the hands of an Internet user in a matter of moments. The user may access the Internet not only at home but at a restaurant, café, on public transport and everywhere in between. Access can take place without apparent detection in the privacy of one’s own home. In addition, research indicates that access to the Internet for information has become the default. The Internet has, in some respects, become a surrogate for memory as has already been suggested.
C. Different Rules Appear to Apply Online
The issue of whether or not there are, or should be, different rules for online behaviour has exercised Internet theorists from the 1990’s to the present day. In some respects, the ease with which information may be accessed seems to suggest that earlier constraints on information access posed by practical and partial obscurity no longer exist. Accessing information on the Internet is more of a free-wheeling exercise aided by the quality of delinearisation of information which allows a user to follow whatever informational trails he or she may wish – from text to audio, to podcast to video, to a Youtube compilation or a learned article of the Social Science Research network.
This lack of constraint suggests to a user that following a query of interest is acceptable, even although it may have been the subject of an earlier prohibition. Why, in an information rich environment such as the Internet, should enquiry be limited? Trial lawyers and Judges have an immediate answer to that proposition, but to one not immersed in the legal culture, to restrict and limit enquiry when the information is so readily available seems counter-intuitive.
Coupled with the private nature of information seeking is the illusion of anonymity given by the Internet. That illusion arises from the fact that an Internet user is rarely identified by name when engaging in a Google search or accessing a news website. The only identification assigned to a user is an Internet protocol number. The relationship between the IP number and the name of the individual is known to the individual’s Internet Service Provider. But unless the Internet user has identified him or herself on a platform, such as Facebook or Twitter, there is a form of anonymity that may engender a sense of immunity from consequences of one’s actions.
- The Online Disinhibition Effect
This sense of immunity is examined in some detail by in an article entitled “The Online Disinhibition Effect” by John Suler.
Suler observes that often people say and do things in cyberspace that they wouldn’t ordinarily say and do in the face to face world. This online disinhibition effect can work in two possible directions. One is benign disinhibition where people share very personal things about themselves, revealing secret emotions, wishes and fears. Toxic disinhibition, on the other hand, involves the use of rude or offensive language, harsh criticisms, anger, hatred, theft and threats. People may visit the dark underworld of the Internet, involving themselves in pornography, crime and violence that they would never explore in the “real world”.
Benign disinhibition may be indicative of an attempt to better understand and develop oneself – a form of “working through” or self-actualisation. On the other hand, toxic disinhibition may simply be a blind catharsis, a form of repetition compulsion and an acting out of unsavoury needs without any personal growth at all.
Suler examines possible causes for online disinhibition and what elements of cyberspace lead to the weakening of psychological barriers that block hidden feelings and needs. He identifies a number of factors. Relevant to this submission is that of “dissociative anonymity”.
An aspect of dissociative anonymity, yet in some respects separate from it, is that of “invisibility”. In text-driven online environments, people can’t see one another. This “invisibility” gives people the courage to go places and do things which they would not otherwise do.
Emily Finch, an author and criminal lawyer studying identity theft in cyberspace, suggests that some people “see their online life as a kind of game with norms and rules that do not apply to everyday living… Once they turn off the computer and return to their daily routine they believe they can leave behind that game and their game identity.”
Suler also observes that within the online environment there is something of a democratisation that takes place with a “minimisation of status and authority”. In the real world authority figures express their status and power in dress, body language and the trappings of their environmental settings. The absence of these together with a lack of the person’s elevated position means that person’s online influence has less effect.
On the Internet everyone has an equal opportunity to voice him or herself. The Internet provides a level playing field and Internet philosophy holds that everyone is an equal and that the purpose of the Internet is to share ideas and resources among peers. This atmosphere and philosophy contribute to the minimisation of authority. Most people, who would normally be reluctant to say what they really think as they stand before an authority figure, are faced online with what is effectively a peer relationship, where the appearances of authority are minimised and people are more willing to speak out and misbehave.
It is submitted that these Internet associated behavioural traits may well provide an insight as to why jurors may feel inclined to ignore judicial direction not to carry out online researches about a case.
F. Internet Democratisation and the Erosion of Authority
The introduction of Web 2.0 and the development of user interactivity has enabled immediate participation within a debate and the ability to share one’s thoughts through the use of blogs, Twitter, Facebook and other forms of social media.
Furthermore, the ability to participate, engage in debate, seek out information and engage with others probably is the greatest opportunity to embark upon a form of participatory democracy. On a global sense, that mirrors the Athenian form of participation and perhaps may even be the first time that the community has had such an opportunity to so engage. The quality of participation is driving many governments towards considering on-line voting, recognising that the Internet enables an opportunity for greater engagement by the community with the political system. The participatory possibilities of the Internet could well mean that, in the future, juries would hear trials on-line rather than being physically present in a court room.
On the other hand, the ability to interact allows many Internet users, especially on social media platforms, to have direct communication with others. This enables the ability to comment, dispute, debate and, sadly, to abuse other Internet users. The online disinhibition effect accounts for this breakdown of restraint in communication. Nevertheless this allows the digital equivalent of a “cat looking at a king” and the normal constraints and deference towards authority figures reduces when dialogue, debate, dispute and commentary become so easy. Although Judges may expect deference to their directions the gradual erosion of respect for authority based upon no other rationale than that of authority alone, especially when a direction may appear to be contrary to rapidly evolving cultural norms and information expectations, means that the weight attributed to a judicial direction may be reduced.
It will by now be apparent that simple reliance upon judicial directions to juries to refrain from accessing information about a case on the Internet may be ill-advised in light of changing cultural attitudes and information expectations. I shall now turn to some examples of juror misconduct in New Zealand and then move on the discuss the issue of take down orders.
- Cases and Examples in New Zealand
The phenomenon of jurors obtaining or receiving information from outside the Courtroom is not unknown in New Zealand. For example, in one reported case, print-outs containing definitions of the “burden of proof” and “beyond reasonable doubt” were found in the jury room. This information had been sourced from the United States and did not correctly state New Zealand law.Examples such as this occur despite observations that have been made about the efficacy of judicial directions and the fact that Judges are able to measure that efficacy in verdicts delivered by juries who, by their verdicts, show that they have understood and followed directions, including those as to prejudice and proper use of evidence.
Although this article takes the decisions in R v Tarapata and Police v Kahia as indicative of the main issues that surround take-down orders, there are a number of other cases that have come before the Courts where take-down orders have been considered. In some cases the orders have been granted, in others they have been refused and I propose briefly to refer to those cases. It should be observed that in the main the same major themes that are present in Tarapata are present in the cases under discussion.
It should be noted that the comments in R v B represent a clear recognition that judicial directions are not a not a completely effective means of preventing jurors engaging in Internet research. Examples of cases where juror research has taken place include M v R and R v Harris where juror research included terms like “burden of proof” and “reasonable doubt.”
In M v R the defendant was convicted in June of 2015 on charges of cultivating cannabis and possession of cannabis for supply. He had previous convictions which had been reported in the New Zealand Herald in 2005. As the result of a communication between a juror and a police prosecutor, there was a suggestion that there had been juror misconduct involving the use of the Internet. This was advanced as a ground of appeal. The Court concluded that there should be an inquiry into the allegation.
The Court considered the provisions of section 76 of the Evidence Act relating to jury deliberations, observing that the section pointed more towards the admissibility of information about jury deliberations rather than prohibiting an inquiry. The allegation in this case was that although the jurors were not satisfied of guilt on the basis of the evidence adduced they conducted their own enquiries and returned a guilty verdict. The result was that Counsel was appointed to conduct an inquiry that was of a specific scope and report back to the Court.
What was significant about this case was that it demonstrates that not only were jurors willing to conduct their own enquiries, but they were also prepared to do so to try and shore up the Crown case against the defendant. If the misconduct were of the nature alleged, not only was there a complete breach of the admonition by the judge to refrain from researching, but clear evidence of partiality and the apparent willingness of the jury to undertake a partisan role.
The case of R v Harris and others was an appeal following conviction on fraud offences. The trial was complex having been brought by the Serious Fraud Office. There were a number of grounds of appeal but the relevant one for the purposes of this discussion arose as a result of a a member of Court staff finding a printout from the Internet site http://www.answers.com containing definitions of the burden of proof and beyond reasonable doubt. These printouts were located between the conclusion of the defence evidence and final addresses and was in clear breach of the trial judge’s introductory remarks in which he specifically directed the jury not to undertake research on the matters arising during trial.
The printout was drawn to the Judge’s attention and he made reference to it in his summing up, pointing out that he was aware that there had been research despite his specific warning. The Judge pointed out in clear terms that the American approach to the matters researched differed from that of New Zealand and that the jury was to take instructions from him and from no other source.
The Court of Appeal observed that the research was in clear breach of the directions of the Judge but observed that the jury room was cleaned out on a daily basis and it was unlikely that the printout would have been available for an extended period of time. The Court noted the strength of the Judge’s direction and observed that juror research has shown that jurors generally follow judicial directions in summing up so the possibility of jurors applying the wrong test was remote.
Whilst the Court was concerned about the possibility of contamination of the jury it was of the view that the possibility was slight.
This case nevertheless demonstrates one of the difficulties about attempting to limit juror research. A take-down or non-publication order may dilute any possible prejudice to the defendant arising from the publication of previous convictions or prior offending. It is impossible to eliminate all possible information that might be the subject of an enquiry. In this case a strong judicial direction had a curative effect, at least as far as the Court of Appeal was concerned. But this remedy would be available only where the juror misconduct was discovered.
An example of the scope of juror research may be illustrated by the case of R v V.V. Reddy. That case involved an online enquiry by jurors about the process by which DNA matches were obtained. During the course of the trial the jury wanted to know how the accused’s DNA was in the possession of the Police and how he was identified as a suspect. Agreed facts were placed before the jury that stated that samples were kept on a National Database, but the process by which they were obtained was not disclosed. A member of Court staff noted that a juror was looking at information about DNA on his phone. The judge interviewed the jurors, two of whom indicated that they were not satisfied with the explanations given and had decided to make their own Internet-based enquiry. The Judge concluded that the problem could not be cured by direction and the trial was abandoned. Once again, that case was one where the prophylactic effect of take-down or non-publication orders would have been inadequate. However, the Judge chose not to impose any sanction upon the enquiring and unrepentant jurors whose breach was quite blatant.
These cases seem to confirm the overseas research that jurors are prepared to carry out their own Internet enquiries in the face of judicial direction and that the consequences of such enquiries can potentially prejudice the outcome of a trial or alternatively require that the trial be aborted.
The case of R v Skelton provides an example of risk management and the importance of judicial proactivity. That was a case which involved issues surrounding the custody of a child who was abducted. The child care issues had been the subject of Family Court proceedings.
An application was made for stay or change of venue on the basis that details of the Family Court proceedings may be recalled by jurors. That was considered to be unlikely, but by the same token the Court considered that the Family Court judgments, which were publicly available on the Internet, should no longer be accessible. The Ministry of Justice was directed to remove the material from the Internet site. The Court recognised that it was impossible to prevent access to the web, but generally jurors would comply with directions not to carry out online research.
Thus the granting of a take-down order in respect of prejudicial webpages is an exercise in risk management and significantly reduces the risk of a compromise to the defendant’s fair trial right. Whilst not acting as a complete answer to juror Internet research it would be nevertheless all the more unlikely for a “googling juror” to access the prejudicial content.
VI. Take Down Orders
Consideration will now move to the nature of takedown orders and then move to a discussion of the decisions in Lyttleton, Tarapata, Kahia and Y v R.
A. What are take down orders?
How does one deal with the quality of persistence of information – with a medium that allows for the continued storage and availability of recorded information. One approach is to make the information difficult to locate. The case of Google Spain SL, Google Inc. v Agencia Española de Protección de Datos (AEPD), Mario Costeja González  which dealt with the so-called “right to be forgotten” in Europe required Google to de-index information so that certain results would not be returned in response to particular queries. Following the decision of the European Court in Google Spain Google put in place a process whereby it would consider requests to de-index information.
De-indexing however only makes the information difficult to locate; it does not remove the information entirely. The public notice advertisement which was the subject of Sr Costeja-Gonzales’ case in Google Spain is still available from the La Vanguardia website in Barcelona.
A take down order requires that a website administrator remove content from its servers or disables access to the content so that it cannot be recovered by a link to its URL. The removal of such content or disabling access to it is a relatively straightforward technological exercise. As will be discussed at a later point in this article, there are significant freedom of expression issues, especially where a take down order is sought against a news media website.
In some respects, take down orders and non-publication orders under s200 of the Criminal Procedure Act 2011 are associated. Non-publication orders usually refer to the non-publication of details of those appearing before the Court and, on occasion, details of the offence and therefore relate to contemporaneous restrictions on publication. A take down order relates to information that may already be in the public domain on an Internet website. In such a situation it may be difficult to make an effective retrospective order, although a takedown order could feasibly be paired with a non-publication order.
The New Zealand Law Commission has discussed the issue of non-publication of prejudicial material in its report on Contempt of Court. Among its recommendations it suggests an association between non-publication orders and take down orders of prejudicial material. The Law Commission proposes a three step process. First, there should be a statutory prohibition upon the reporting of an arrested person’s previous convictions for the purpose of preserving the right to a fair trial.
Secondly, there should be provision for an order postponing publication of other information if it appears necessary to avoid a real risk of prejudice to a fair trial. Such an order could be made at any time after arrest and before the trial has been completed, but it could not extend beyond the completion of proceedings.
Thirdly, and importantly for the present discussion, there should be provisions authorizing the court to make an order that an online content host take down or disable public access to any specific information covered by the statutory prohibition in any suppression order made under the proposals above.
The Law Commission recommendations are reflective of the problems posed by the Digital Paradigm. They recognise the fact that the quality of persistence of information coupled with the ready availability of Internet based content of an historical nature can potentially be prejudicial to the fair trial rights of the accused. Whereas a contemporaneous publication could conceivably amount to a form of publication contempt, historical material may well have been published legitimately but, in light of changed circumstances, have taken on a prejudicial element. In this respect the proposals by the Law Commission are prophylactic in nature.
As far at the Courts were concerned the issue of take down orders had not come to any prominence until the case of Lyttleton v R which I shall discuss in greater detail below. However, that case established that the Court had jurisdiction to make a take down order as part of its obligation to protect the fair trial rights of an accused person. Since that case there have been a number of instances where take down orders have been considered by the New Zealand Courts.
B. Some New Zealand Take Down Cases
In the three or so years between Lyttleton and Y v R, there have been a number of cases involving the removal of material from the Internet. These have not been widely publicised. They illustrate that the Courts are prepared to grapple with the problems posed by the availability of Internet based information. The issue of juror compliance with directions was considered by Ellis J in R v Lyttle.
- R v Lyttle
Mr Lyttle was charged with murder in May 2011. His trial was due to commence late in 2017. In 2012, he was convicted of 5 charges of obtaining money by deception. There was an application by the Crown to have these convictions admitted as propensity evidence. The convictions were ruled inadmissible. They had little probative value and would have been prejudicial to Mr Lyttle.
When the convictions were entered there was publicity including an article published on the Stuff website. It was one of the first “hits” on a search using the string “David Lyttle Builder” on both Google and Bing.
The Crown sought an order that the article and links to it be taken down during the course of the trial. The application was supported by the defence and opposed by Fairfax.
Ellis J considered the decision of the Court of Appeal in Lyttelton and the observations regarding the remoteness of a juror locating potentially prejudicial articles on the Internet and the comment that there was an expectation that jurors would comply with directions not to conduct their own enquiries. However, Ellis J drew attention to the earlier case of R v B where William Young P and Robertson J suggested it “must be commonplace” for jurors to make internet inquiries such as googling the accused, despite judicial directions. They went on to say:
 A specific direction to jurors not to “Google” the defendant may put the possibility of doing so into the mind of a juror who might not otherwise have thought of it. There may be scope for debate as to whether it is best for a judge to square up to the detail of the prejudicial material or to deal with the topic with generalities. The reality is that there is no simple and foolproof way for a trial judge to address the availability on the internet of prejudicial material about a defendant.
Ellis J also referred to the District Court case of R v Scott where the Judge ordered that articles relating to Mr Scott’s previous convictions for sexual offending be taken down during his trial on sexual violation charges.  The Court of Appeal commented favourably upon the take-down order saying that, together with standard directions, it would be safe to mitigate any potential risk.
Ellis J then went on to consider recent research on Internet use by jurors. She observed that the Court of Appeal in Lyttelton relied on 1998 research by the Law Commission but had not noted that the Law Commission in 2014, in its Issues paper on Contempt in Modern New Zealand, expressed the view that the 1998 finding underrepresented the extent to which jurors use the Internet to find information during a trial because of the increased use of the Internet in the twenty-first century. Indeed, Ellis J observed – correctly in my view – that “there is simply no meaningful comparison to be drawn between internet use in 1998 and its ubiquity now.”
In allowing the take-down order, Ellis J considered the following matters to be relevant:
- The article was over 4 years old and there was no contemporary public interest in it. In addition, the take-down was of limited duration and the value attached to freedom of expression was lower than in other cases. This contrasts with the suggestion in other cases that the historical nature of the material mitigates against a take-down.
- There was only one article which would be the subject of the order and again the freedom of expression was very limited. The practicalities of take-down were straightforward.
- The matters in the article had been ruled inadmissible and were prejudicial, thus the content had been the subject of evaluation.
- Notwithstanding that faith in the jury system was predicated on the fact that jurors would comply with judicial directions, “there is empirical evidence (which was not before the Courts in Lyttelton) that strongly suggests that that is not always the case in relation to internet searches.”
- The unusual circumstances of the case and the way in which the defendant was implicated may pique jury interest and prompt what the Judge referred to as “defendant googling”.
Arguments against the order were that it could be presumed that jurors would comply with directions, although the Judge was of the view that this could not be absolute. Coverage of the trial would not link to the earlier article although that would not address the “googling juror” problem. She did pause at the argument that the removal of the article would not remove it from a Google search in that the snippet may still be present, but it was acknowledged by Fairfax’s counsel that removal of the article would make it difficult, even for a determined juror, to locate.
Accordingly, a take-down order was made.
Lyttle was decided on 4 October 2017. Tarapata was argued on 30 October 2017 and Moore J released his reasons for his decision on 10 December 2017. Counsel in Tarapata referred to Lyttle which was noted by Moore J. However, he placed faith in the power of the judicial direction as a means of mitigating the risk that jurors may carry out their own enquiries.
Other cases decided, however, assume that there is such a risk.
- R v Tranter
In R v Tranter, decided on 4 November 2015 there was an application for an order that Fairfax Media, courtsofnewzealand.govt.nz, newzeelend.wordpress.com, odt.co.nz, and 3news.co.nz were to remove any article from the internet that named the defendant, David Stanley Tranter, together with details of any conviction, sentence or court appearance involving him. The order was to subsist until the disposition of Mr Tranter’s trial.
The application was dealt with under urgency and subsequently Fairfax applied to have the order cancelled. This was opposed by the defendant and the Crown.
This was a case where it appears that specific articles were not identified. Fairfax had used best endeavours to comply with the order. The articles complained of were historic and related to the defendant. The only details of any evaluation of content were that it was considered that the content should not appropriately be viewed by jurors or potential jurors.
Gendall J referred to Lyttelton which was relied upon by Fairfax, especially the comments about juror compliance with directions. On the other hand it was argued that if the material did remain online the case was more serious than Lyttelton and if the order was rescinded the fair trial rights of the defendant would be adversely affected.
The Judge agreed, which suggests again that there had been some evaluation of the content and it was observed that the material could be prejudicial although the decision does not say why or how.
Once the trial was over and the defendant had been found guilty the take-down order came to an end.
- R v Scott
R v Scott was a decision of the District Court on 28 June 2017.  The case involved a joint application by the Crown and the defence for a take-down order in respect of information relating to the defendant who was to undergo trial for sexual violation. It was proposed that material be taken down from Google as well for a period equivalent to the length of the trial.
It was observed that although the defendant had the benefit of name suppression up until trial, that would not assist him because his identity would be revealed to jurors who may use his name as the basis of an Internet search.
Fairfax Media opposed the application. It conceded that the District Court had the inherent power to regulate its process and to make such an order.
Judge Butler referred to the competing points of view on whether jurors would seek out their own information – Lyttelton v R on the one hand where it was held that the Court must proceed on the assumption that jurors will follow those directions and resist the temptation to make their own enquiries on the Internet; and the other, the observation in R v B
“Jury research has established that jurors often make their own inquiries despite judicial directions not to do so. Internet inquiries, perhaps just in the form of “googling” the defendant, must be commonplace. This means that publicity about a defendant can no longer be assumed to be of only transitory significance.”
A specific direction to jurors not to “google” the defendant may put the possibility of doing so into the mind of a juror who might not otherwise have thought of it. There may be scope for debate as to whether it is best for a Judge to square up to the detail of the prejudicial material or to deal with the topic with generalities. The reality is that there is no simple and foolproof way for a trial Judge to address the availability on the Internet of prejudicial material about the defendant.
Thus, Judges could not be confident that jurors would not obey instructions to refrain from making their own enquiries.
Fairfax suggested that the risk was not as great as it may seem.
- What if there was an appeal – what would happen to the short term order.
- Floodgates – that there would be an increase of such applications in jury trials.
- Take-down assumed that jurors would breach their oath.
- The order would not prevent a dedicated investigator seeking the information out.
The Court ordered that the order apply to Fairfax and to Google.
It should be noted that in Scott there does not appear to be any identification of the content by way of reference to URLs or other specific identifiers, nor does there appear to have been any evaluation of the content to determine whether or not it would be prejudicial. From a risk analysis perspective, this step is an important one. It can be inferred from the fact that the application was a joint one by Crown and defence that there was a recognition of the prejudicial nature of the content. Finally, there is no express discussion of the tension between the risk of prejudice to a fair trial and the freedom of expression and the Press. It seems to have been assumed that the fair trial interests of the defendant would be prejudiced to such a degree that a take-down order was justified.
- McMahon v Fairfax Media
The case of McMahon v Fairfax Media was unusual in that it centered upon a suppression order.  The accused was charged with burglary. He unsuccessfully sought name suppression in the District Court, but in the High Court Lang J ordered that details of the offending be suppressed. It was noted that as long as the details were suppressed there was no reason to believe that members of the public would have cause to access the Internet.
Subsequently, Fairfax published a report on the “Stuff” website. The report detailed the charges against Mr McMahon without naming him. An application was made for a take-down order in respect of the article. Courtney J noted that the purpose of the suppression order was to ensure that details of the offending remained suppressed. She was of the view that that there was a distinct risk that members of the public would realise what had happened and connect Mr McMahon with the offending reported on Stuff. A take-down order was made.
Interestingly enough, no comment was made about what appears to have been a breach of the suppression order. However, the primary focus of the case seems to be that of ensuring the integrity of the order. There was no expression of concern about possible juror enquiry, although that was a matter which concerned Lang J and an unstated concern that there could well be a downstream effect of connection of the unusual offending with the defendant.
Tranter, Lyttle and Scott demonstrate a willingness on the part of some Courts to accept that there is a risk that jurors will conduct their own enquiries on the Internet and that prejudicial material should be taken down.
However, there have been four decisions, starting with Lyttleton which suggest something of a resistance to the making of take down orders and a preference for a reliance upon judicial directions to deal with the issue of the “Googling Juror”
- Lyttelton, Tarapata, Y v R and Kahia
- The Decision in Lyttelton
The case of Lyttelton is an unusual one.  It is important because it is the first case in which appellate consideration was given to the issue of Internet take-down orders.
Mr Lyttelton had been convicted of a number of violence charges to which he pleaded guilty and was sentenced to a term of imprisonment. There was media coverage of the case and articles about it were published in hard copy and online. Mr Lyttelton served his sentence and then appealed his convictions and a retrial was ordered. An order was made prohibiting publication of any of the proceedings including the result in the news media, on the Internet or any other publicly available database until disposition of the retrial.
In March 2015, with support from the Crown, Mr Lyttelton applied to the High Court for orders that the historic online articles about his case be taken down, arguing that the availability of those articles would be prejudicial to his retrial. Orders were made by Lang J that the article in question be taken down from the various websites. The media organisations involved removed the material but sought recission of Lang J’s order and following a hearing Lang J revoked his earlier order. The appeal by Mr Lyttelton was against that revocation.
Lang J made five findings.
First, the articles were historic in nature and did not automatically appear when an Internet user went to a media home page. It would be necessary to search the website or employ a search engine such as Google to locate the content.
Secondly, it was unlikely that members of the jury pool would have a residual recollection of the publicity accompanying Mr Lyttelton’s case in 2009 or 2010 and the suppression orders would ensure that the media would not report on the matter.
Thirdly, Mr Lyttelton’s identity would be unknown to jurors until they were empanelled. At that time they would receive directions from the Judge.
Fourthly, those directions would be firm and clear that the jury was only able to consider the evidence placed before the Court and that they were not to conduct their own research.
Finally, the risk of juror enquiry on the Internet would only arise after jury selection and by that time directions would be given. Lang J was of the view that the Court had to proceed on the assumption that jurors would follow those directions and resist the temptation to make their own enquiry on the Internet.
There was considerable discussion in the judgment about the nature of the case and the proper appeal track, along with the question of whether or not the Court had jurisdiction to make the orders. Once the nature of the appeal was settled the Court was able to turn to the substantive appeal.
For the media, submissions from Fairfax were considered. It was argued that there was no sufficient risk to Mr Lyttelton’s fair trial rights to warrant reinstatement of the order and that the matters addressed by Lang J were compelling. Further it was argued that research carried out by the Law Commission in 1999 found that jurors were rarely aware of pre-trial publicity, that they made an effort to focus on the evidence before the Court and there was no evidence that they were affected by media coverage. The media also gave undertakings that the material would be restored to its original URLs and that media coverage of the trial would not link to those articles.
The Court considered that there was no real prejudice demonstrated by Mr Lyttelton and that Lang J’s approach was persuasive. The case was not in the public consciousness and a juror would have to actively search to locate the material. This was considered to be a remote possibility and did not justify the “drastic measure of removing all coverage of his previous trial from their respective online platforms.”
The Court observed that this was not a high profile trial like the retrials of Bain or Lundy where pretrial publicity was seen as inevitable and irremediable in terms of juror exposure to it. The absence of contemporary media interest was a factor the Court considered.
Importantly, the Court emphasised that it agreed with the observations of Lang J that if directions are given by a trial judge jurors can be expected to comply with them, and there was no suggestion that this was unlikely to occur.
- The Decision in Tarapata
In 2015, Mr Tarapata was tried and convicted on two charges of murder. His convictions were overturned on appeal and a new trial was ordered. The re-trial was set for late 2017. The 2015 trial had been extensively reported in the media. Those reports were available on the Internet.
Shortly before Mr Tarapata’s retrial, counsel applied for extensive suppression orders, including orders requiring various media organisations to forthwith remove from their websites references to Mr Tarapata’s first trial, including all content and electronic links.
The basis for the application was that if any members of the jury at the retrial undertook an internet search, they would discover details regarding Mr Tarapata’s first trial, which was considered highly prejudicial and would compromise his fair trial rights.
Justice Moore made without notice interim orders but directed that the various media organisations be served, reserving leave for them to apply to revoke the orders. Following service upon the media organisation, they filed notices of opposition and affidavit evidence. During the course of the trial, the Judge heard argument from all parties and determined that the interim orders that he had made should not be continued and they were rescinded. He reserved his reasons which were delivered on 19 December 2017.
1. The Factual Background
On 19 July 2014, Mr Tarapata entered a pawn shop on Great South Road in Takanini. He attacked the two victims with a knife and stabbed them both to death. He then fled the scene with his partner and finally drove to a Police Station in Huntly, where he gave himself up.
It appeared that his motive for killing the men was that he believed they were having a sexual relationship with his partner. This belief was wholly unfounded and was driven by an intense paranoia and jealousy caused by psychosis and the defence of insanity was advanced. It was accepted that Mr Tarapata was delusional and psychotic.
The Crown’s case at the first trial was that Mr Tarapata’s psychosis was caused by an external influence, mainly his chronic use of methamphetamine. The defence position was that Mr Tarapata’s use of methamphetamine actually exacerbated an underlying organic psychiatric illness, known as schizophrenia, so that at the time of the killing, he was suffering from a disease of the mind.
As far as moral wrongfulness was considered, it was claimed by the defence that Mr Tarapata was driven by religious delusions and believed that God, as a higher moral authority, had directed him to kill the two men. The Crown on the other hand argued that although Mr Tarapata was psychotic and delusional, he knew exactly what he was doing and intended the consequences. He was simply a jealous person, even though he was psychotic and delusional.
The jury rejected Mr Tarapata’s defence and found him guilty and he appealed to the Court of Appeal. The Court of Appeal allowed the appeal and ordered a retrial. There were criticisms of the way in which the psychiatric evidence had been adduced and, in particular, the fact that much of the psychiatric opinion had not been based on primary evidence produced during the trial and tested by cross-examination. There were criticisms made in relation to what the Court determined was the unfounded opinion of the Crown’s psychiatric expert on the role and influence of methamphetamine on the question of a disease of the mind.
At the second trial, Mr Tarapata was represented by new counsel. There was considerable co-operation and consultation between the Crown and the defence. In contrast to the first trial, the Crown agreed not to lead any evidence relating to Mr Tarapata’s methamphetamine use and the defence and the Crown had agreed that Mr Tarapata was suffering from a disease of the mind at the time of the killings. Thus, the second trial proceeded on a much narrower basis. The only issue was whether or not the defence had proved, on the balance of probabilities, that at the time of the killings, Mr Tarapata did not know that his acts were morally wrong, having regard to the commonly accepted standards of right or wrong.
The first trial, as has been noted, was extensively covered in the media and there was a significant amount of emphasis in the media upon the evidence relating to his methamphetamine use. Approximately 10 days before the second trial, the defence counsel filed a detailed Memorandum drawing the Court’s attention to the body of the material available on the internet from Mr Tarapata’s first trial, including a transcript of the Judge’s full sentencing decision.
It was argued that it would be in the interests of justice for this material to be removed to reduce the risk of jury members undertaking internet searches on their own, discovering that there had, in fact, been a previous trial but that Mr Tarapata had been convicted. Accordingly, take-down orders were sought. On the eve of the trial, defence counsel filed a further Memorandum, observing that since the earlier discussions, further investigations had been undertaken and screen-shots were provided from what was described as a quick and simple search of the internet using Mr Tarapata’s name.
The next day the trial commenced and the jury was empanelled. The Judge made remarks to the jury, emphasising the importance of bringing an open mind to the task of judging and directed the jury in stronger terms than usual not to undertake any enquiries of their own. Special mention was made of a prohibition upon undertaking internet and social media searches. The Judge’s directions followed recommendations of the Law Commission in its discussion paper “Reforming the Law of Contempt of Court”.
The Judge also made interim orders, directing that the various media organisations, whose websites contained various references to Mr Tarapata’s 2015 trial, were to remove all content and electronic links relating to him and his trial. The interim orders that were made are important because they set the scene for what was to follow and give important context to the approach of the media organisations and their opposition, as well as the way in which the Judge subsequently dealt with the matter. The interim orders read as follows:
“I am prepared to make the interim orders sought on a without notice basis. I direct that they are to continue until further order of the Court. Accordingly, I direct as follows:
The listed media outlets below and any other media outlets served with this order forthwith remove from all websites within their direct and indirect control all references to the 2015 trial of the defendant, Zarn Tarapata, in this matter, including all content of electronic links to related material:
Television New Zealand
The operators of the search engines known as Google, Yahoo and Bing and any other operator of a search engine served with this order forthwith disable any and every link between a search term using the words “Zarn” and “Tarapata” (individually or in combination) and a website to the effect that users will not be directed to any websites that report any aspect of the defendant’s trial in 2015 in this matter, including subsequent sentencing; and
These orders will continue until the conclusion of the trial, or such other times the Court may direct.
If any party wishes to oppose the continuation of the orders, it is to file and serve a notice of opposition, setting in full the grounds of such opposition. That notice is to be filed and served within 48 hours of the making of these orders.
In the event of opposition, it will be necessary to convene a hearing and, if necessary, hear evidence. I emphasise that the present interim orders are in order to preserve Mr Tarapata’s fair trial rights in the context of my assessment of the prejudicial material contained in the media reports I have been referred to. Given the limited time within I have been obliged to deal with this matter, that assessment has not been as full or as considered as I might have wished. It is possible that following any hearing my provisional view might change.
I direct that the orders be served electronically to the email addresses set out above”.
Once the orders had been served, references to Mr Tarapata and his first trial were removed from the websites. Google searches using Mr Tarapata’s name led to links which did not allow access to or the downloading of material.
However, in response to the opportunity to be heard, the news media organisations opposed the making of the orders. The Judge summarised the 13 objections by the news media, which fell into a number of generalised categories.
First, the argument was advanced that the take-down orders were an unreasonable limitation on the media’s right to freedom of expression. It had not been demonstrated that there was specific prejudice or extreme circumstances from which the Court should infer prejudice to Mr Tarapata’s fair trial rights. No risk of prejudice had been demonstrated by the continuing availability of the historical articles relating to the first trial online. Thus, the take-down orders were not necessary to ensure fair trial rights were preserved.
Secondly, the objection was raised that the jury could be expected to follow judicial directions.
Thirdly, the media also submitted that there was prejudice to the media in terms of compliance with the take-down orders, which involved extensive effort by the news media to identify likely material that should be taken down. Furthermore, there was no guarantee that all of the material could be located, given the published articles they had copied, scraped or commented upon by third parties on other websites and on social media.
The Judge, in coming to his conclusion, first considered that the risk of prejudice to fair trial rights had to be a real one and that that risk would remain despite the strong direction given to the jury that they should not undertake any enquires of their own.
Furthermore, in assessing the risk of prejudice of fair trial rights, the Judge made reference to the fact that this was not a notorious case and that it was unlikely that jurors would bring any residual memory or knowledge of the earlier trial of Mr Tarapata to the first trial, thus meaning that their curiosity would not be piqued so that they would undertake an internet search.
On several occasions throughout his Judgment, the Judge emphasised the importance of residual memory or absence of public consciousness and used this to bolster his conclusion that it would be unlikely in that event that jurors would undertake an internet search. He considered that there was no real or substantial risk that a determined or irresponsible juror might undertake his or her own internet enquiries.
2. Compliance Issues
The Judge’s orders were broadly expressed, as can be seen from the text above. Essentially, what the order required was for media organisations to search their content management systems for key terms, such as Mr Tarapata’s name. This might not necessarily identify every piece of published content which might forward in the scope of the take-down order.
In addition, there was no control over what results Google or other search engines could display in response to a particular search and even when content was taken down, there may be snippets of information scraped from a particular article which might appear on the search results.
There were added compliance difficulties in terms of eliminating from the video content of news reports any specific content that related to Mr Tarapata’s trial. Essentially, an entire news report would have to be taken down to eliminate access to a particular part of the video record – what the Judge referred to as deleting the historical record.
This led to the conclusion by the Judge that compliance with the orders was oppressive to the media organisations and that considerable resources were required to be re-directed from core business functions.
But the real problem, as far as this case was concerned, is articulated at paragraph  and following.
It was pointed out by Google New Zealand that the interim orders did not identify the content with sufficient specificity to enable a search engine provider, such as Google, to determine what content was required to be removed. The search engine provider requires a full and complete URL in order to prevent the web pages of those URLs from being returned as search results. Without URLs being specified, a search engine provider can only speculate as to which particular web pages may be covered by the orders. Thus, search engine providers had difficulty in identifying specific content, as it was not part of their function to carry out a search for particular content and then determine whether or not the search engine would identify it.
In this commentator’s view, the absence of specific URLs that could identify content raise a number of issues:
(i) As has been pointed out, it is difficult for search engine providers, such as Google, to de-index specific content without a URL.
(ii) It is, perhaps, a tribute to the news media organisations that they were able to search their content management systems to locate potentially prejudicial material and take it down.
(iii) However, the absence of specific items meant that it was not possible for the Court to carry out an evaluation of the material to determine whether or not it was sufficiently prejudicial to warrant a take-down order. Without being able to consider the various items on a case by case basis, a proper evaluation and balancing of fair trial rights against freedom of expression and freedom of the press could not have been undertaken.
It is this writer’s view that this, in and of itself, could justify Moore J in rescinding the take-down orders. However, as has been noted, there were other factors surrounding the outcome of the Tarapata decision.
- The Decision in Y v R
The case of Y v R was an application for take-down orders in respect of 9 identified online articles published local news websites.
The applicant was charged with the murder of M in Northland in March 2016. His trial was due to commence in July 2018. Initially the take-down application was very wide, but, by the time it came on for hearing, it was refined to 11 identified articles. There was a further refinement during the hearing so that 9 articles were the subject of the application.
The argument on behalf of the applicant was that if jurors were to carry out an Internet search based on the name of the applicant, the address where the murder was alleged to have taken place or of some identified names, it would become clear that:
- The applicant was a member of a gang
- He had faced and still was to face other charges
- He was on the run from the Police for some time
- At that time he was considered armed and dangerous
- He was involved in another shooting where a person had died but in respect of which there was no causative link between the shooting and the death.
It was argued that, given the nature of the case, there was a real risk that a juror might become curious and seek out information about the applicant on the Internet and that risk existed irrespective of the direction that a Judge may make.
The media argued that the material was historic, implying that there was little currency in the material and that there was little public interest in the content. That meant that a juror would have to undertake a search for it and any risk that he or she might do so would be mitigated by appropriate directions by the trial judge. There were also freedom of expression issues raised by the case which the Court had to take into account.
Wylie J identified the tension between freedom of expression and fair trial rights that were raised by the take-down application which underpinned the fact that the inherent jurisdiction to make a take-down order should not be exercised lightly, but only where the risk to fair trial rights was a real one, rather than a remote possibility. A real risk is one that is more than speculative in that there must be a likelihood of prejudice to the administration of justice.
In a comprehensive decision which contained citations not only of New Zealand but also Australian authority, Wylie J assessed the issue of reality of risk under a number of heads. First, he observed that there was nothing remarkable about extensive pre-trial publicity. However, it was possible for any potential prejudice to be mitigated by the trial Judge. In some respects, this reflects the reality of partial obscurity.
Secondly, and associated with the first point, the Judge could direct jurors to take into account only the evidence that was heard in Court. This was an important consideration for Wylie J. There was a stated assumption that jurors would comply with directions and their legal obligations in considering whether there was a sufficiently real risk to warrant a take-down order. Similarly, the judge referred to Australian authority which holds that the court must be of the view that jurors may be inclined to seek access to material on the Internet, despite directions by the Judge.
Thirdly, in determining whether there is a real risk of prejudice, the Court must consider the likelihood of material coming to the attention of a juror who wants to seek it out. Wylie J drew a distinction between a high profile case where there was relatively recent material with that which was less prominent in the public eye or where the material was historic. Wylie J referred to the decision of the Court of Appeal in Lyttelton v R. In that case, the Court was of the view that there was an absence of real prejudice in that the articles in question were old and were not immediately available on the websites of the media organisations. It was noted that for a juror to search for them, a degree of proactivity would be required, using either the defendant’s name or that of the victims which the Court considered was a remote possibility. The Court considered that the case was not in the public consciousness and was unlikely to be the subject of significant pre-trial media coverage in a way that would make juror exposure inevitable.
There were other matters which may be taken into account in determining the risk of prejudice, including the likelihood of innocent non-compliance, the oppressive nature of compliance and the alteration of the historical record, if it was unlikely that the material may not be restored to the Internet. On the other hand, commercial considerations would take second place to fair trial considerations.
In coming to his conclusion that the take-down order was not justified, Wylie J observed that the 9 identified items were historic in nature, having been published on the Internet in February, March and April of 2016. Historic articles, he concluded, were rarely searched although they could be found. Specific search terms were required such as the names of individuals, or addresses. If that information was not held or was unknown, it was unlikely that a person would be able to identify historic articles, even using specific searches.
Another significant factor was that the likelihood of a person carrying out a search before being selected as a juror was remote given that such a person, if summoned, would have no advance knowledge of the cases set for trial, nor details which might enable a search to be carried out. If there is to be a risk, it is once the jury is empaneled and Wylie J placed considerable store on the efficacy of careful and detailed directions. If pre-trial publicity was a concern, directions could be extensive including advice to jurors that carrying out independent enquiry would be a breach of court orders, could amount to contempt and render them liable to penalties. In addition, jurors could be directed to bring a breach by another juror to the attention of the trial judge. In the case before him, Wylie J concluded that there was no evidential basis for assuming that jurors would ignore such directions.
Innocent non-compliance was not an issue in this case, given the way in which the articles had been refined and media representatives confirmed that the mechanics of take-down were not difficult. It was observed that there were difficulties with search engines and with those who may have republished potentially prejudicial material meaning that prejudicial material may remain on the Internet. In the same way that material was easy to take-down, it could easily be restored, thus meaning that the historical record would not be lost.
In the final analysis there was nothing which justified interfering with the freedom of expression by making a take-down order.
The approach in this case is similar to that in Tarapata. In Y, however, the content was identified and was evaluated. As it stood, the content was potentially prejudicial. Like Tarapata, there was a strong emphasis upon the efficacy of judicial directions and the associated matter that a casual enquiry by a juror was unlikely to uncover the material. Thus, there would have to be a degree of determination and a wilful flouting of a judicial direction to construct the necessary search strings that would locate the material.
D. The Decision in Kahia
Moore J, who decided Tarapata, also decided the takedown application in Police v Kahia. In December 2014, Kahia was convicted of murder. His conviction was overturned by the Court of Appeal the following year. The matter was set down for re-trial which commenced on 14 May 2018.
At the first trial, the identity of the person who caused the fatal injuries to the deceased was the key issue and would be the key issue at the re-trial. At the first trial, the defence was that the defendant was not the killer, who may have been any one of a number of other people that he was with that evening.
When the matter came up for re-trial, it became clear that the defence would be argued with greater particularity, in that it would be argued that it was not Mr Kahia who inflicted the fatal wounds but another specified person. Thus, Crown witnesses would be examined a good deal more comprehensively about the nature of their prior relationship and their relationship with previous co-defendants.
An application was made for take-down orders of publicity about the 2014 trial. It was supported by an affidavit. This listed the available on-line material relating to Mr Kahia’s first trial, convictions and sentence. The deponent, Ms Opetaia, undertook a Google search of the words “Hendrix Kahia”. She also searched other word combinations, including “Hendrix Kahia murder”, “Gang Related Taupo Murder” and “Wiremu Birch.” As a result, 17 articles were located. The URLs were provided. Four of the articles related to the offending itself and the charges brought against Mr Kahia and his co-defendants. A further four related to evidence given during the first trial. Three related to Mr Kahia’s conviction. One result was a YouTube interview with one of the witnesses. The other articles reported upon the imposition of a sentence of life imprisonment upon Mr Kahia.
Moore J observed that take-down orders had been considered by senior Courts on three previous occasions. He discussed the case of Lyttleton v R, his own decision of R v Tarapata, as well as the decision of Ellis J in R v Lyttle.
Moore J considered that an analysis of the case law revealed the following principles:
- A Court must be careful to balance the right to freedom of expression with a right to a fair trial, conscious that the extent to which the interests will be engaged will depend upon the context of the application.
- Because the right to a fair trial is absolute, the context may require an order impinging upon the media’s right to freedom of expression by ordering the removal of new stories until the conclusion of the trial.
- The threshold question is whether there is a real risk that the applicant’s fair trial rights will be adversely affected if the material remain available. The test is whether there is a real or substantial risk a determined and irresponsible juror might undertake internet enquiries (citing Tarapata).
- In resolving that question, the following factors will be relevant:
(a) The level of notoriety, that is whether pre-trial publicity will be inevitable and simply irredeemable in terms of jury exposure to published material.
(b) The likelihood that despite compliance with the take-down order material out of the control of the media will, nevertheless, remain available on line, which is in contravention of the order.
(c) Whether interim suppression orders will have the practical effect of permanently removing material from the public historical record and the imposition on the media in terms of the cost of compliance.
Moore J noted the conflicting High Court authority on the question of whether in assessing the real risk a jury member might undertake their own enquiries. He noted the Court should consider empirical evidence.
He referred to the decision of Ellis J in Lyttle, noting that there was in fact empirical evidence that was not always the case that jurors would obey jury directions and that the presumption that they would do is not absolute. However, by contrast in Tarapata, Moore J concluded it must be expected that a juror would adhere to very firm and tailored directions and that was a fundamental principle upon which the criminal justice process was founded. Although it may be likely that disobedient jurors would undertake their own enquiries despite a judicial direction, there has to be judicial satisfaction that the real risk threshold had been crossed.
The Judge observed that pre-trial publicity necessarily carries a risk of prejudice to fair trial rights but the question was whether or not there was a real risk to those fair trial rights.
The Judge discussed the issue of whether jurors may be expected to adhere to judicial directions. Even with the presence of adverse publicity and widespread media interest, there was authority which suggested that, notwithstanding this, jurors would focus on the evidence before them, as the material most immediately and recently to hand for their assessment and it was not to be assumed that jurors would ignore judicial directions to put to one side matters they may have heard outside the court. Experience showed that jurors were responsive to judicial directions and tended to be more robust than defence counsel often assume.  Once Moore J had disposed of the issue of whether or not jurors might carry out their own enquiries in the face of strong judicial directions prohibiting such action, the outcome was quite predictable. If jurors were not going to “google”, what possible prejudice might there be?
There are some observations that should be made about Moore J’s approach in Kahia. Clearly the decision in Kahia reinforces Moore J’s approach in Tarapata. There are some elements of nuance present in this decision that were not available in the earlier one, including a consideration of the conflicting approach of Ellis J in Lyttle. The Judge goes to some pains to distinguish that case primarily upon the nature of the information that was the subject of the take-down order and the amount of it. Interestingly enough, he impliedly accepts that in Lyttle the real risk threshold had been crossed but it is suggested that a significant element was the prejudicial nature of the material.
What is of concern is the dismissive approach to the empirical evidence that is available. Some (but not a great amount) of the research material available was presented to the Court. The material submitted from the United States was in the nature of news articles than any careful empirical or academic studies, a number of which are readily available. Nevertheless, it seems that Moore J has preferred to follow the established rubric articulated in Lyttleton, rather than shift his position on his reliance upon jury directions.
The question must be whether there may be a concern on the part of the Judge that to acknowledge that judicial directions may not be followed may strike at the heart of the jury system itself. If jurors are going to Google in breach of jury directions, are they more likely to ignore other directions about more fundamental aspects of the jury role? Perhaps Moore J is concerned that by acknowledging the fallibility of judicial directions in this area that he perceives a potential erosion of the jury trial process.
There are a couple of other matters that arise from this decision. One is that there seems to be a suggestion in the judgment that the take-down orders will remove prejudicial material. This, of course, ignores the exponential dissemination quality of information on the internet. What is sought to be achieved is a dilution of prejudice rather than a complete removal of it. A court needs to understand how search engines work and the way in which removed material, coupled with the indexing in the page ranking system used by Google, will mean that other incidents of prejudicial material are less likely to appear in a Google search.
In addition it is suggested with respect that the reliance upon the “historical record” argument is a red herring. A media content management system will easily be able to replace an article that has been taken-down and it is for that reason that it is suggested that take-down orders should be finite in duration to enable media organisations to replace the material once the order has expired.
- The Themes Arising from the Cases
The four cases discussed have a number of common factors to them which can be expressed as general themes that underly the decisions.
- Juror Compliance with Judicial Directions
The first theme is of the issue of compliance by jurors with judicial directions not to engage in private research and especially online research. This theme seems to resemble an article of faith by judges and is seen as the prophylactic against jurors conducting their own research or carrying out their own enquiries.
There may, however, be a deeper concern. As I have observed, if a Judge is prepared to accept that jurors may ignore a prohibition against online research, perhaps implicitly there is a lack of confidence that jurors will follow other directions given during the course of the trial or, even more importantly, during the summing up.
The concern can be addressed in two ways. The first relates to modern reliance on technology discussed earlier in this article. The tendency for reliance upon devices is to obtain immediate access to information. Thus the Internet may be accessed to obtain details about a particular location from a maps application or Google Earth, about one of the lawyers involved in the case, from a Google search or a review of Facebook, or about an accused from a Google search that might turn up an article containing details of previous convictions.
However, despite the concerns already expressed about the democratisation of information and the erosion of authority, it is suggested that judicial authority will add weight to directions on matters of law, such as the burden and standard of proof and the elements of an offence. In addition, visual aids such as question trails give added reinforcement to judicial directions on matters of law, thus demonstrating a difference in the quality of information that may be sought by way of a Google search.
- The Quality of Digital Information
The second theme relates to the nature of information on the Internet. Contemporaneity and notoriety of reporting on cases is contrasted with what is referred to as “historical material” which is unlikely to remain in the residual memory of a juror. This is akin to what has been referred to as the “fade factor” and involves consideration of pre-Internet assumptions about the nature of information, wrapped up in the terms “practical” and “partial obscurity”. These assumptions are challenged by the qualities of digital information in the Digital Paradigm, especially the availability of information by use of a search engine which is one of the most common navigation tools on the Internet.
The problem that must be recognised is that our assumptions about information and the validity of partial and practical obscurity must change and reflect the fact that we live in a new information paradigm where “historical” information is readily available and appears fresh and in the same format as when it was first made available. This is an issue that will increase as more and more information is digitised by default and “hard copy” becomes the exception.
- Freedom of Expression vs Fair Trial Rights
The third theme – a very important one – lies in the tension between freedom of expression and the freedom of the press to report Court proceedings with the right of an accused person to a fair trial.  In the cases for fair trial rights to trump freedom of expression, there must be a real possibility of prejudice as opposed to speculation. The “real possibility” argument has been addressed in the cases by reliance upon juror compliance with directions and the unlikelihood of jurors breaching those directions and locating the prejudicial information.
With the exception of Y v R, very little, if any, consideration has been given to the specific identification of potentially prejudicial content and an evaluation of that content for prejudice. Once the assumptions involving compliance with directions and the determination of a juror to locate content have been put to one side, there must be an identification of the individual items of content and an evaluation of that content for prejudice.
D. Media Compliance
A fourth theme involves the issue of compliance by media organisations with identifying content, taking it down and restoring it with associated considerations of possible interference with the historical record. Given the sophisticated content management systems employed by mainstream media organisations and the ease with which information may be located and relocated in the digital space, the “historical record” will receive but minimal interference given that a take down order will be of limited duration, after which the media organisation will be at liberty to restore the information once the order has expired. If there is a concern that might not happen and that a media organisation may overlook the opportunity to restore the historical record, it is suggested that this is an issue that reflects more on the information management systems of the media organisation than as a justification for refusing a take down order. Furthermore, the sanctity of the historical record cannot be seen as inviolate, given the developing concept of the right to be forgotten and the new provisions in the EU General Data Protection Regulation which replaces the former “right to be forgotten” with a right of erasure.
The fifth theme involves dealing with the quality of searchability of content using a search engine and whether or not search engine platforms such as Google and Bing may comply with deindexing requests. It should be emphasised that in Tarapata what Google was required to do went beyond the de-indexing of identified content and required Google to evaluate content which it was reluctant to do. I shall address the issue of de-indexing below.
- Efficacy of Take Down Orders
The final theme which is addressed in the conclusion is that of the efficacy of take-down orders. Information persistence is a quality of Internet information together with that of exponential dissemination. There may well be cases where information about a case has spread beyond news media websites and may be located in unusual and little-known internet locations. If the information is available on Facebook, Twitter, Instagram and similar platforms, it can be identified and consideration given to its removal. But it must be emphasised that in the context of the Internet, a take-down order is not going to eliminate prejudice. It is only going to dilute its prejudicial impact.
- Freedom of Expression
A take-down order against a media outlet impinges upon the freedom of the press and the freedom of expression enjoyed by news media organisations as guaranteed by s 14 of the New Zealand Bill of Rights Act 1990.
The freedom of expression must be balanced against other affirmed rights and freedoms. This applies not only to those contained in the New Zealand Bill of Rights Act, but in other areas, such as minimum standards of criminal procedure and fundamental principles of law, such as the protection and promotion of the free and impartial administration of justice.
Parallels arise between take-down orders of prejudicial material pre-trial and elements of what the Law Commission refers to as “publication contempt”. Indeed in its recent report on the law of contempt, the Law Commission recommended a reform of the law to enable the “take-down” of online content, such as the previous convictions or publication of other material that may prejudice a fair trial.
The tension between freedom of expression and fair trial rights arises frequently. The freedom of expression right protects open justice and the reporting of court proceedings. At the same time, the fair trial rights of a defendant have received recognition and may require an exception to the open justice principle.
Section 5 of the New Zealand Bill of Rights Act 1990 confirms that a freedom such as that of free expression, ought to be restricted only so far as necessary to protect a countervailing right or interest, in that the freedoms contained in the Act are subject only to such reasonable limitations as are prescribed by law and are demonstrably justified in a democratic society.
Within the context of fair trial rights, the issue is whether the particular interference with the administration of justice is so serious as to override the freedom of expression. Using the law relating to contempt as an example, especially that of “publication contempt,” the objective – protection of a fair trial or the administration of justice – must be sufficiently important to justify limiting the freedom of expression. The limitation imposed must interfere with the freedom as little as possible.
In the case of Gisborne Herald Co Ltd v Solicitor General it was held that where there was a real risk that the publication of an article would prejudice a fair trial, freedom of speech rights and the importance of a fair trial and the proper administration of justice could be accommodated by deferring publication until after the trial.
The overriding importance of a fair trial in the context of the administration of justice was exemplified in the case of R v B (CA459/06) where the tension arose in the shape of an application for a non-publication order and in respect of which Baragwanath J stated that “a fair trial trumps all”.
It is suggested that the making of a take-down order amounts to a justified limitation of the news media’s rights and is supported by authority. Emphasis is placed upon the word “limitation”. A take-down order should be limited only to the duration of the trial and no longer – such time as is necessary to ensure protection of the defendant’s fair trial right and to ensure that a self-informing juror is unable to find the articles the subject of the order. Once the trial is over, access to the material can be re-enabled.
Similarly limitations consistent with ensuring a level of free expression that aligns with the administration of justice and the right to a fair trial could be addressed by restricting the scope of the order solely to material that details the fact that a defendant has previous convictions or material from which that may be inferred and which would interfere with the presumption of innocence and associated fair trial rights.
The obligation on counsel to carefully consider and assess the prejudicial content is considerable. If evidence is going to be adduced, for example, of a defendant’s gang connection, it is unlikely that a take-down order would be made in respect of such information. This is because although publication of such information could be prejudicial, it would be raised within the context of the trial. On the other hand, if an earlier article included details of a defendant’s trial on a similar matter, including details of his previous convictions and criminal conduct, and there was to be no propensity application, such information could be prejudicial.
In some respects, the quality of information persistence that characterises Internet based information, together with that of searchability, places earlier publications of potentially prejudicial material into a grey area between the protection of a fair trial by means of a take-down order of prejudicial material and “publication contempt” which, in pre-Internet times was concerned more with inflammatory material that was published in mainstream media while a trial was pending or a case was sub judice.
X. The Law Commission, Publication Contempt and Take Down Orders
The Law Commission in its recent discussion paper and report on Contempt was of the view that the current common law rules surrounding publication contempt do not contain any “bright line” directions and are unclear. The scope of publication contempt is considered to be uncertain which may have a chilling effect on public discussion. On the other hand, an overly robust approach could be taken which could compromise fair trial rights.
The clear principle emerging from the Law Commission report on contempt is the primacy of the importance of the fair trial as a justification for interfering with the freedom of expression.
One of the problems identified by the Law Commission was when those reporting events get it wrong and compromise fair trial rights. The problem is that it is difficult to remedy such a problem after the fact. If it becomes apparent that there is some form of publication prejudice, trials may have to be abandoned at considerable expense and inconvenience, not to mention speedy trial rights and an erosion of public confidence in the ability of the Courts to deliver justice. Further difficulties arise where a breach of fair trial rights is uncovered after conviction and sentence, revealing an unsafe conviction which must be set aside and a retrial ordered.
The Law Commission expressed some concern about the “real risk test”. It was suggested that there should be a separation between assessing whether there is a risk and determining whether or not that risk can be mitigated.
Finally, the Law Commission recognised the way in which information use and expectations have changed in the Internet age. Anyone may publish information or post images and video at any time. This strengthens the need for certainty and clarity in the law surrounding contempt.
The Law Commission proposals for prohibitions upon the publication of an arrested person’s previous conviction, for other information that should be suppressed to protect fair trial rights and an associated power to make take down orders have been discussed above:
It was recognised by the Law Commission that publication of previous convictions before trial would normally amount to common law contempt. It recognised that there was potential harm that arose from disclosure of this information and that the freedom of expression gave way to fair trial rights. The temporary limitation of a suppression order recognises the importance of fair trial rights. Although the proposal by the Law Commission is more restrictive than the common law, it clarifies the limitations that may be placed on pre-trial publicity and would deter publication of material that might jeopardise a fair trial.
The proposal for take-down orders against an online content host, requiring it to take-down specific material if it was necessary to protect fair trial rights would, if enacted, remove the take-down regime from Judge-made law to statute. A take-down order was recognised as a temporary measure for a particular purpose.
It was recognised that take-down orders would not be a perfect or complete solution, nor would they deter or prevent the determined internet user but the Law Commission concluded that they could go some way towards minimising the impact of an offending publication.
In summary therefore, publication pre-trial of previous convictions would be prohibited, although the prohibition should be kept under review. The postponement of publication of other material could be prohibited to avoid a real risk of prejudice to a fair trial and take-down orders could be made in respect of these two classes of information.
The essence of the Law Commission proposals recognises the importance of a fair trial and that displaces freedom of expression. However, the displacement is only within the context of the trial and the limitation of the freedom of expression is clearly circumscribed. Following conclusion of the trial, publication prohibition would no longer apply and material the subject of a take-down order could be restored.
- Identification and Evaluation of Content
It will be clear, especially from the discussion about freedom of expression issues, that an application for a take-down order should be made only in the most obvious cases of prejudice. This means that counsel should evaluate the content of the material carefully and be sure that the prejudice relates to the particular defendant and the issues that are before the Court. The precise identification of content does not appear from the decision of Moore J in Tarapata, although 9 specific articles were identified by Wylie J in Y v R. In this writer’s view, one can only demonstrate a justified limitation on the freedom of expression by precisely identifying material and aligning it with the issues at trial to establish prejudice.
Once prejudicial content has been identified, the likelihood of retrieval will have to be demonstrated. The first step will be to develop search parameters and undertake careful searches for information based on those parameters. A record of the search parameters must be kept because the methodology of the process will have to be explained to the Court.
The record should include not only the search parameters but the date and time of the search. It would be advisable to take “screen dumps” of the search results to validate the results of searches. The results acquired by the searches may result in different rankings for the same material. A careful record of the rankings must be kept and the searches should be repeated over a period of days or even weeks to determine consistency of rankings and return. This exercise will identify possibly prejudicial material and establish a pattern of rankings that will provide an indication of the likelihood of retrieval using the various search parameters identified.
The search process and methodology, together with the results, should be recorded in an affidavit together with copies of the articles and their Universal Resource Locators (URL) and the search result screen dumps exhibited.
Alternatively, if the trial budget allows it, the services of a reputation management consultant could be obtained to carry out an extensive search and analysis of all potentially prejudicial material present on Internet platforms. Such an exercise might go well beyond mainstream media websites and include posts on Facebook, Instagram, Twitter and other social media platforms.
Because the substantive case is a criminal trial, any application that is made will be ancillary to the criminal proceedings. The reality is that although the principal parties will be the Crown and the Defendant, the online content hosts, be they news media websites or otherwise, will have to be served. Thus, in addition to an application for a take-down order, there should be an application for directions as to service of the take-down application upon the interested parties.
A comprehensive memorandum in support of the application clearly identifying the nature of the prejudice and the necessity for orders should be prepared and filed.
The order sought should only be as wide as necessary to dilute the prejudicial effect of the content. The maximum time frame for a take-down order would be for the duration of the trial and until the jury returned a verdict. Then access to the material could be restored. Thus, the order sought would be along the lines that the identified web-page content and associated URLs be removed from the website, or access thereto be disabled from the date of commencement of the trial until such time as the jury at the trial returns with a verdict.
- Desirability of an Order and Meeting the Google Problem
It was clear from the decision in Tarapata that Google LLC, based in California, did not consider itself subject to the order that was made by Moore J.  In the view of this writer, part of the problem lay in the fact that the scope of the order was too wide, and went beyond the deindexing process that Google has put in place.
The difficulty with the order in Tarapata was that Google was treated as a content provider, rather than as the provider of a means of locating content on the Internet. The roles may seem similar but in fact are substantially different. A mainstream media organisation such as the NZ Herald or Stuff are online content hosts. They have sophisticated content management systems which can be used to locate the content of a particular article. Google, on the other hand, scours the Internet for content and rather than preserve it (other than in a cache) indexes it and links to the particular source of the content. Other than a brief description or “snippet”, Google does not make the content available.
The primary source of reference for a Google search is a URL which enables a linking from the search result to the webpage where the content is located. The Google deindexing process means that the search results – the URL and hypertext link – do not appear in any searches for the content. Google removes those URLs from its search index.
Thus, in seeking a take-down order, the specific content should be identified by URL, not only for the purposes of prejudice evaluation, but also because it will be of assistance in a subsequent approach to Google.
Once the application is successful and an order is obtained the next step is to approach Google to deindex the content. This approach is necessary because a juror may conduct a search which returns a result and a Google snippet of the content, but trying to obtain the content by hyperlink would be unsuccessful. The prejudice is that the juror would be aware that at some time content of interest was available. Thus the ability to obtain such a result may be constrained by deindexing.
Following the “right to be forgotten” decision in Google Spain instituted a process whereby requests can be made to deindex content. The process may be initiated at Google’s Legal Removal Requests page. A copy of the Court Order for temporary removal of the content will have a persuasive effect upon Google’s decision to deindex. The process suggested would obviate the necessity for a deindexing application to be made which would be a complex, time consuming and expensive process involving, to start with, service upon Google in the United States.
- The Future of Take-down Orders
The necessity for take-down orders will probably increase as Internet use continues and the availability of information online becomes an accepted way of informing oneself. The trend, at least in recent politics, which suggests that citizens are not simply prepared to accept the say so of an authoritarian figure means that jurors are more likely to go online to augment or verify the information that they hear in Court. Short of sequestering the jury, a restriction on the availability of prejudicial information would seem to be the only solution.
However, as has been suggested in this article, this is a remedy which should be used sparingly and only in the clearest cases. The Courts will be careful to scrutinise applications for take-down orders, mindful of the tension between the freedom of expression and fair trial rights. Nevertheless, it should be observed that although this article has directed attention primarily at mainstream media websites, the rise of the “citizen journalist”, the ready availability of Internet based publication platforms and their ease of use, and the development of private commentators on justice matters means that prejudicial material in these “new media” locations may need to be considered. One aspect of the matter that must govern whether or not a Court will interfere with this information is its ease of location utilising a search engine.
The final observation that should be made is that any prejudice that may be occasioned by the existence of online material may only be diluted and not totally eliminated by its removal and de-indexing. The solution of a take-down order may only mitigate or dilute prejudice, but it may nevertheless go part of the way towards ensuring a fair trial and addressing the problem of the Googling Juror.
 LLB (Auckland) MJur (Auckland) PhD (Auckland); Judge of the District Court (Acting Warrant); formerly Director, New Zealand Centre for ICT Law and Part-time lecturer in Law and Information Technology, Faculty of Law, University of Auckland. I acknowledge the assistance and inspiration provided by Justin Harder, Adam Holland and Katherine Maxwell together with Rosemary Tobin. I also acknowledge the assistance of Sarah Watt who offered a number of helpful suggestions on an earlier as well as the final draft. Sections of this article have appeared in another form in a discussion of injunctions and publication restraints in R Tobin and D Harvey Entertainment and Media Law in New Zealand (Thomson Reuters, Wellington, 2017) at p 89 et seq.
 David Harvey “The Googling Juror: The Fate of the Jury Trial in the Digital Paradigm”  NZ L Rev 203.
 For a detailed discussion see David Harvey Collisions in the Digital Paradigm: Law and Rulemaking in the Internet Age (Hart Publishing, Oxford 2017) especially at Ch 2 p 16 et seq.
 Harvey above n 2 at p 226.
 Other terms to describe the nature of the order sought include “gagging order”, although the context of the granting of a prior restraint injunction to prevent publicity of defamatory, confidential or private information occupies a different space in the control of publication of information.
 Sam Hurley “Google thumbs its nose at New Zealand Courts” 23 May 2018 NZ Herald https://www.nzherald.co.nz/business/news/article.cfm?c_id=3&objectid=12056284 (last accessed 24 May 2018).
Sam Hurley “Considerable Concern at Google’s Unwillingness to follow Court Orders” 23 May 2018 NZ Herald https://www.nzherald.co.nz/business/news/article.cfm?c_id=3&objectid=12057169 (last accessed 24 May 2018).
 R v Bailey High Court, Auckland CRI 2007-085-007842, 23 April 2010 per Winkelmann J.
 Iti v R  NZCA 492.
 Bailey, above n 7 see especially paragraphs  et seq.
 R v Reddy  3 NZLR 666 at .
 Reddy above n 10.
 The term “practical obscurity” was used in the case of US Department of Justice v Reporters Committee for Freedom of the Press 489 US 749 (1989).
 For a recent discussion of practical obscurity in the context of the availability of personal information in on-line court records see Jane Bailey and Jacquelyn Burkell “Revisiting the Open Court Principle in an Era of Online Publication: Questioning Presumptive Public Access to Parties’ and Witnesses’ Personal Information” (2017) FIMS Publication 159 p. 168-169 http://ir.ib.uwo.ca/fimspub/159 (last accessed 29 April 2017).
 Gisborne Herald v Solicitor-General  3 NZLR 563 at 567. For a discussion of the law relating to contempt of court and publication contempt, see Rosemary Tobin and David Harvey New Zealand Media and Entertainment Law (Thomson Reuters, Wellington, 2017) at Chapter 6. See also New Zealand Law Commission Reforming the Law of Contempt: A Modern Statute R140 (New Zealand Law Commission, Wellington, 2017) http://www.lawcom.govt.nz/our-projects/contempt-court?id=1417 (last accessed 27 August 2017).
 Viktor Meyer-Schonberger Delete: The Virtue of Forgetting in the Digital Age (Oxford University Press, Oxford 2010).
 For a full discussion of the qualities of digital information see David Harvey Collisions above n 3 at Ch 2 and especially p 22 et seq. In developing a taxonomy of qualities each is broadly classified as environmental, technical and user associated.
 Harvey above n 3 at p37.
 Steps to limit or restrict the operation of search engines, as was the case in the “right to be forgotten” case of Google Spain SL, Google Inc. v Agencia Española de Protección de Datos (AEPD), Mario Costeja González (2014) ECJ Case-131/12 http://curia.europa.eu/juris/liste.jsf?num=C-131/12&language=EN have a significant and detrimental effect upon the overall utility of the Internet.
 Ronald N Kostoff “Expanded Information Retrieval Using Full Text Searching” (2010) 36 J Information Science 104.
 Filippo Menczer “Complementing Search Engines with Online Web Mining Agents” (2003) 35 Decision Support Systems 195; Bailey and Burkell, above n 13 at p. 170 http://ir.ib.uwo.ca/fimspub/159 (last accessed 29 April 2017).
 See Y v R  NZHC 489.
 A Gibson, and others “The Internet in New Zealand 2013” (Institute of Culture, Discourse & Communication, AUT University Auckland 2013) https://icdc.aut.ac.nz/__data/assets/pdf_file/0005/73445/wipnz2013final.pdf (last accessed 10 August 2017).
 C Crothers, and others “The Internet in New Zealand 2015” (Institute of Culture, Discourse & Communication, AUT University Auckland 2016) p. i. https://workresearch.aut.ac.nz/__data/assets/pdf_file/0003/71328/WIPNZ-Report-060515.pdf (last accessed 10 August 2017).
 A Techatassanasoontorn and others Karimikia “World Internet Project New Zealand – Internet in New Zealand in 2017 (New Zealand Work Research Institute, Auckland 2018) especially at 5 https://workresearch.aut.ac.nz/__data/assets/pdf_file/0009/174915/Internet-in-NZWIP-2017.pdf (last accessed 7 September 2018) – see also InternetNZ The State of the Internet 2017 (InternetNZ, Wellington 2018) https://internetnz.nz/sites/default/files/SOTI%20FINAL.pdf at p. 4 – 8. (last accessed 7 September 2018)
 Cheryl Thomas “Are Juries Fair?” (Ministry of Justice Research Series 1/10, February 2010) https://www.justice.gov.uk/downloads/publications/research-and-analysis/moj-research/are-juries-fair-research.pdf (last accessed 10 August 2010)
 Above n. 25 at p. viii.
 For other examples of Internet misuse by jurors see New Zealand Law Commission Reforming the Law of Contempt: A Modern Statute above n. 14 at para. 4.14 p. 73.
 Cheryl Thomas “Avoiding the Perfect Storm of Juror Contempt”  Crim LR 483
 Thaddeus Hoffmeister “Google, Gadgets and Guilt: Juror Misconduct in the Digital Age” (2012) 83 U Colo L Rev 409.
 Hoffmeister above n 29 at 414 – 415 (footnotes omitted). For other examples of juror misconduct see above n 2.
 Hon Antoinette Plogstedt “E-Jurors: A View from the Bench” (2013) 61 Cleveland St L Rev 597 at Part V.
 Robbie Manhas “Responding to Independent Juror Research in the Internet Age: Positive Rules, Negative Rules and Outside Mechanisms” (2014) 115 Michigan L R 809.
 Nicole L Waters and Paula Hannaford-Agor “Jurors 24/7: The Impact of New Media of Jurors, Public Perceptions of the Jury System and the American Criminal Justice System.” National Center for State Courts Center for Jury Studies http://www.ncsc-jurystudies.org/What-We-Do/~/media/Microsites/Files/CJS/What%20We%20Do/Jurors_%2024-7_REV011512.ashx (last accessed 8 September 2018).
 Paula Hannaford-Agor, David B Rottman and Nicole L Waters “Juror and Jury Use of New Media: A Baseline Exploration” (National Centre for State Courts, Williamsburg, 2012) http://www.ncsc-jurystudies.org/~/media/Microsites/Files/CJS/New%20Media%20Study/NCSC-Harvard-005-Juror-and-Jury-Use-of-New-Media-Final.ashx
 Above n 34 at p 8.
 Harvey above n 2 at 208 – 209.
 For examples see Gareth S Lacy: Untangling the Web: How Court should Respond to Juries using the Internet for Research” (2011) 1 Reynolds Court and Media Law Journal 169 at 173-176. http://issuu.com/rnccm/docs/reynolds_courts_and_media_law_journal_vol_1_issue_/43?mode=embed&viewMode=magazine. (last accessed 8 September 2018). See also the examples cited in Harvey above n 2 footnotes 26 – 38.
 Hon Amy J St. Eve and Michael Zuckerman “Ensuring an Impartial Jury in the Age of Social Media” (2012) 11 Duke L & Tech Rev 1; Hon Amy J St Eve, Hon Charles P Burns and Michael Zuckerman “More from the #Jury Box: The Latest on Juries and Social; Media” (2014) 12 Duke L & Tech Rev 64.
 Hannaford-Agor, Rottman and Waters above n 34.
 Viktor Meyer-Schonberger above n 15.
 For example see David R Johnson and David Post “Law and Borders – The Rise of Law in Cyberspace” (1996) 48 Stanford LR 1367;
Jack Goldsmith “Against Cyberanarchy” (1998) Univ Chicago LR 1199;
Judge Frank H Easterbrook “Cyberspace and the Law of the Horse” (1996) Univ Chicago Legal Forum 207;
Laurence Lessig Code and Other Laws of Cyberspace (Basic Books, New York 1999);
John Perry Barlow “A Cyberspace Independence Declaration” available at https://www.eff.org/cyberspace-independence (Last accessed 6 September 2018);
A.M. Froomkin “The Internet as a Source of Regulatory Arbitrage” in B Kahin & C. Nesson (eds) Borders in Cyberspace (MIT Press, Boston, 1997);
Henry Perrit “The Internet as a Threat to Sovereignty? Thoughts on the Internet’s Role in Strengthening National and Global Governance” (1998) 5 Ind J Global Legal Stud 423. For critical comment see J Goldsmith “The Internet and the Abiding Significance of Territorial Sovereignty (1998) 5 Ind J Global Legal Stud 472;
Molly Land “Towards an International Law of the Internet” (2013) 54 Harvard Int LJ;
Lawrence B Solum “Models of Internet Governance” in Lee A Bygrave and Jon Bing (eds) Internet Governance: Infrastructure and Institutions (Oxford University Press, Oxford 2009) at 48.
 For a discussion of the quality of delinearisation of information see David Harvey Collisions above note 3 at chapter 2.
 John Suler “The Online Disinhibition Effect” (2004) 7 Jnl of Cyberpsychology and Behaviour 321.
 Above n 43 at 321.
 Above n 43 at 321.
 Above n 43 at 322.
 Above n 43 at 322.
 Cited in Suler above n 43 at 323. In some respects this lends weight to the perception that different rules apply online.
 Above n 43 at 324.
 Above n 43 at 324. For further discussion, see below under the heading “Internet Democratisation and the Erosion of Authority”
 Above n 43 at 324.
 This phenomenon is not restricted to Internet based research. Despite “practical obscurity” issues, jurors have visited the scene of a crime (R v Gillespie CA 227/88, 7 February 1989), conducted experiments to work out how long it takes for a car engine to cool down (R v Taka  2 NZLR 129 (CA)). or how much heroin could be concealed in shoes (R v Sangraksa CA 503/96, 3 July 1997) and enquired of chemists about the availability and price of ephedrine. (R v Bates  1 NZLR 326 (CA)). None of these cases resulted in a prosecution for juror contempt.
 R v Harris CA 121/06, 27 September 2006
 R v Tarapata  NZHC 3209
 Police v Kahia  NZHC 1023
 R v B (CA 459/06)  NZCA 130;  1 NZLR 293 at  – .
“Jury research has established that jurors often make their own inquiries despite judicial directions not to do so. Internet inquiries, perhaps just in the form of “googling” the defendant, must be commonplace. This means that publicity about a defendant can no longer be assumed to be of only transitory significance.”
“The reality is that there is no simple and foolproof way for a trial judge to address the availability on the internet of prejudicial material about the defendant.” 
 M v R  NZCA 37.
 R v Harris above n 53.
 R v Harris above n. 53.
 R v V.V. Reddy  NZDC 10437.
 R v Skelton HC Hamilton CRI-2006-019-006530 9 July 2008
  NZCA 279;  2 NZLR 21;  1 PRNZ 1.
 R v Tarapata above n. 54
 Police v Kahia above n. 55
 Y v R above n. 21
 Above n. 18.
 New Zealand Law Commission above n 14.
 There should be a requirement that the pre-trial or trial court to keep the prohibition under review and authorise the court to lift, extend or vary the prohibition as necessary in any particular case. The prohibition should apply from the time a person is arrested and only where the person is arrested for an offence for which he or she is liable to be tried by a jury (a category 3 or 4 offence).
 New Zealand Law Commission above n 14 at p.7.
 Lyttleton v R above n. 62.
 Above n. 62
 Above n. 21
 R v Lyttle  NZHC 2426.
 R v B (CA 459/06) above n 56 at . For further on R v B see below in the discussion about New Zealand cases of juror misconduct.
 R v Scott  NZDC 13939.
 Scott v R  NZCA 357 at  and .
 R v Lyttle above n 73 at .
 R v Lyttle above n 73 at (d).
 Tarapata above n 54 at .
 Tarapata above n 54 at .
 R v Tranter  NZHC 2727.
 R v Scott above n 75.
 Lyttelton v R above n 62.
 R v B above n 56.
 McMahon v Fairfax Media  NZHC 1812.
 Above n. 62
 Lyttelton v R above n 62.
 R v Lyttelton  NZHC 763 (HC).
 Lyttelton v R above n 62 at para .
 Above n 54
 Tarapata above n 54 at para .
 Tarapata above n 54 at para .
 Above n 54 at paras  and 
 Y v R above n. 21
 Y v R above n. 21.
 Y v R, above n 21 at paras  – ; Gisborne Herald Co Ltd. v Solicitor-General above n. 74 at 567.
 Solicitor-General v W & H Specialist Publications Ltd  3 NZLR 12 (HC) at .
 Fairfax Digital Australia & New Zealand Pty Ltd v Ibrahim  NSWCCA 125 at  and .
 Lyttelton v R, above n 62.
 Lyttelton v R, above n 62 at  – .
 Y v R, above n 21 at .
 Police v Kahia above n. 55.
 There have in fact been others – see R v Tranter above n. 81; McMahon v Fairfax Media above n. 85 and Y v R above n. 11. There is also the District Court case of R v Scott above n. 75.
 Above n. 62.
 Above n. 54.
 Above n. 73.
 Above n 54. The decision of Wylie J in Y v R above n. 21 articulates this test. The Court must be satisfied that a real risk exists, despite a direction to the jury not to undertake enquiries of their own.
 R v Lyttle above n 73.
 See R v Rickards HC Auckland CRI-2005-063-1122, 28 November 2005
 For details of the material that was available as at 2014 see Harvey above n 2.
 I shall discuss the important issue of freedom of expression below.
 General Data Protection Regulation, Article 17.
 Solicitor-General v Radio New Zealand Ltd  1 NZLR 48 (HC) at 59, and see comments in R v Chignall & Walker [1990-1992] 1 NZBORR 179.
 Solicitor-General v Radio New Zealand Ltd, above n 113; Duff v Communicado Ltd  2 NZLR 89 (HC).
 New Zealand Law Commission Reforming the Law of Contempt of Court: A Modern Statute above n 14 p 36 – 51. For a summary of the proposals for take-down orders see p 7 R 3.
 Siemer v Solicitor-General  NZSC 68;  3 NZLR 441 at  – ; L v R  NZCA 279;  2 NZLR 21 at .
 Duff v Communicado Ltd above n 114 at 100.
 Solicitor- General v Radio New Zealand Ltd above n 114.
 Gisborne Herald Co Ltd v Solicitor General above n 14 at 575.
 R v B above n 56.
 R v B above n 56 at [2 ].
 A similar approach was adopted in Gisborne Herald Co Ltd v Solicitor General above n 14.
 NZ Law Commission above n 14.
 NZ Law Commission above n 14 pp. 46 – 47.
 Above n 14 p. 48 para 2.55
 See the discussion above under the heading of VI Take Down Orders.
 Above n 14 p. 48 para 2.58
 Siemer v Solicitor-General above n 116.
 Tarapata above n 54.
 Y v R Above n 21.
 Tarapata above n 54.
 Tarapata above n 54.
 Google Spain SL, Google Inc. v Agencia Española de Protección de Datos (AEPD), Mario Costeja González above n.66.
 https://support.google.com/legal/answer/3110420?visit_id=1-636293565525935582-2797058458&rd=1 (last accessed 3 May 2017).