Section 127 of the Communications Act 2003, once one of the more obscure provisions of the cybercrime world, has had a good workout lately[1]. Famously, Paul Chambers, delayed at Doncaster Airport and frustrated at possibly not getting to see his girlfriend, was accused and convicted of sending ‘by means of a public electronic communications network a message or other matter that is grossly offensive or of an indecent, obscene or menacing character‘ (s 127(a)) because he had sent a humorous and frustrated tweet saying: ”Crap! Robin Hood Airport is closed. You’ve got a week and a bit to get your shit together otherwise I am blowing the airport sky high!!’.
After a long period of civil liberties campaigning, support from celebrities and comedians and sustained outrage among the Twitterati sub nom #TwitterJokeTrial, on the third attempt, an appeal court[2] saw sense and conceded that ”a message which does not create fear or apprehension in those to whom it is communicated, or who may reasonably be expected to see it, falls outside this provision [of the 2003 Act]’. In other words, a joke issued to the world as well as the very people it might offend and indeed identified with the real name of the author, is clearly not meant to be taken seriously and thus is neither menace nor threat for any reasonable person.
Meanwhile, however, s 127 prosecutions continue or are urged on by a public increasingly fed up with an increasing variety of online racist bullies, trolls, stalkers etc. For example, in early 2012 racially motivated tweets posted relating to the footballer Stan Collymore[3] were also prosecuted under this legislation. In late September 2012, Neil Swinburne, 29, was also prosecuted under s 127 for putting up a page apparently glorifying an alleged murderer of a policeman http://www.guardian.co.uk/uk/2012/sep/20/police-killings-arrest-cregan-facebook Meanwhile on the very same day, the DPP, Keir Starmer, was driven by the rise of s 127 prosecutions and apparent mounting calls for its catch-all use in any case of disturbing content on social media, to declare that he would be issuing guidance on social media prosecutions. Asked to consider whether to prosecute Daniel Thomas, who had made trollish and homophobic tweets about the Olympic diver Tom Daley,[4] the DPP indicated that s 127 should not be seen as a carte blanche for prosecuting content which, however upsetting to some, would normally fall with guarantees of freedom of expression in a democratic society. In particular he quoted the seminal ECHR case of Handyside v UK[5] which says that freedom of expression includes the right to say things that ‘offend, shock or disturb the state or any sector of the population‘ before concluding that prosecution in the Daley case would not be appropriate.
We therefore now await the promised guidance. But as the editor of this journal has cogently pointed out, guidance on its own is simply not good enough. The law itself must have a quality of predictability and certainty, otherwise the rule of law is in jeopardy. This is especially true of a criminal provision which restricts freedom of expression. Why then, we might ask, is s 127 drafted so widely? Partly because, although it appears to be a modern post-Internet provision , its direct and very close antecedents actually date from long before the Internet era and even before Handyside. These antecedents include the Post Office (Amendment) Act 1935 (and two subsequent PO Acts) – which dealt with messages sent by post and telephone – the British Telecoms Act 1981 and the Telecommunications Act 1984, s 43. Section 127 of the 2003 Act basically repeats the 1984 Act provision wholesale, itself almost a word for word repetition of these earlier Acts. The 1984 Act was notable for changing the scope to apply to any ‘public telecommunication system’ (a necessity following the deregulation of the state monopoly telephone network in 1981). This phrase itself was taken, it seems, from EC telecoms law, and causes problems of its own, debated in Chambers (see further below).
The legislative history of s 127 is usefully narrated in DPP v Collins[6] (at [6]), a case involving a man who made repeated telephone calls to his local MP’s office asking for him to do something about the ‘black bastards’ – and similar even more unpleasant terms. He was charged with sending ‘grossly offensive’ messages under s 127. On appeal to the House of Lords, the charge against Collins was upheld but what is particularly interesting for present purposes lies in LJ Bingham’s analysis at [7] of what the purpose of s 127 is.
It is crucial here, parenthetically, to note that s 127 is hardly a lone legislative bulwark against Internet trolls and harassers. There is an abundance of law to deal with what the CPS calls ‘communication offences’ or alternately ‘public order offences’.[7] Indeed so much law is available to charge social media content that appears to offends the (or a) public that it is often hard to find out from journalistic coverage exactly what legislation is being used, though the majority of cases involving Twitter do seem to use s 127. For example, in another recent controversial case, 19-year-old Azhar Ahmed was convicted of ‘racially aggravated public order offences’ after he posted an angry Facebook status update about the reporting of the latest British Army fatalities in Afghanistan. It appears s 127 of the Communications Act 2003 was used to charge in this case, but such cases could however also fall under the Public Order Act 1986 (POA), s 18 . By contrast, when Liam Stacey, a 21-year-old biology undergraduate, made ‘racially aggravated comments’ (ie tweets) about footballer Fabrice Muamba, he was apparently charged under the Crime and Disorder Act 1998[8] before being sentenced to 56 days in jail. Other sections of the POA are also potentially relevant to social media comments: notably s 4A which deals inter alia with ‘threatening, abusive or offensive words’ which cause ‘intentional harassment, alarm or distress’.
The Protection from Harassment Act 1997 (PHA), which operates slightly differently in England and Scotland, has also been used successfully on several occasions to charge trolls who send repeated upsetting or vile messages to users on sites like Facebook and Twitter. The PHA prescribes that any two ‘acts’ which form a course of harassing conduct can be charged as a crime. These provisions were recently used by, eg, Nicola Brookes, who, to great publicity, won a Norwich Pharmacal order against Facebook in order to reveal the true names of, and start proceedings against, her online trolls under the 1997 Act. An Adjournent Debate in Parliament on 17 September 2012 noted not only these but also the possibilities of using the Public Order Acts and the Computer Misuse Act. Private civil damages can also be obtained both under the PHA and by common laws like libel.
Finally there also exists (albeit in England and Wales only), the Malicious Communications Act 1988 which, though little mentioned in recent social media cases, still exists and is of significance as a contrast to s 127.[9] Again a pre-Internet statute, its original target seems to have been poison pen letters[10], although it was updated in 2001 to apply to ‘electronic communications’ – ‘oral or otherwise[11]‘ But the speciality of the 1988 Act is that it is intended to apply only to one-to-one exchanges and not to one-to-many broadcasting. It explicitly prescribes that the communication must be ‘sent to another person’. So the 1988 Act would not, it seems, apply to Paul Chambers telling the world at large about his frustration at Doncaster Airport, nor (say) the racist bullies who left tweets for Fabrice Muamba and Stan Collymore but also distributed them to the world.
Thus, returning to LJ Bingham in Collins, he observed (at [7]) that the existence of the 1988 Act and thus he deduced the purpose of s 127 was ‘not to protect people against receipt of unsolicited messages which they may find seriously objectionable‘. Instead, it is ‘to prohibit the use of a service provided and funded by the public for the benefit of the public for the transmission of communications which contravene the basic standards of our society‘.
A lot can be unpicked from this interesting dictum. LJ Bingham is here partly making the distinction raised above between an offence of sending criminal content to one person (intended to be covered by the Malicious Communications Act) and broadcasting it one-to-many (which he sees as within the scope of s 127). This was relevant for the Collins case as the people whom the communications might reasonably have ‘grossly offended’ were seen as, generally, people of ethnic background – who had not personally received the calls – as opposed to the MP’s staff, who had. Accordingly it was held that under s 127 it was not necessary for the actual recipients of the communication to have been grossly offended. – an objective rather than subjective test.
But the judge was also clearly aware that the preceding ancestor statutes to s 127 originated from a time of state monopoly services over post and phone when it would have been impossible for an ordinary person using only ordinary postal or phone facilities to broadcast their views. This was true even at the time of the Telecommunications Act 1984 – which is the first of the statutes in question to use the phrase ‘public telecommunication system’ – since the Internet could not be regarded as publicly available in the UK before the early 1990s[12]. What LJ Bingham seems to home in on, then, is that s 127, as originally formulated back in 1935, was not just about public morality but also about not wasting public money on transmitting material which was unpleasant to the public. As such, the words used extend to categories of speech which are wider than the bare minimum acceptable in democratic society. Section 127 is thus readable as the heir to provisions which once reasonably criminalised even what would now be permitted speech using the Handyside test. The proximate reason s 127 is thus more restrictive than the ordinary criminal law on speech not via electronic means is crucially not because a distinction is made between online and offline speech, but because it involved a public facility such as the postal service, or later, pre-deregulation British telecoms services[13].
Yet this justification for the broad extent of s 127 no longer exists in the cases surveyed above, such as those involving Chambers, Swinburne and Thomas. Twitter is a private service run on private servers. So is Facebook. People use the Internet to access it, yes, but this at most now involves the use (sometimes) of the facilities of former public utilities. In essence, however, communication on social networks is privately funded and devoid of any concern for public funding priorities.
In Chambers, this point was indirectly raised with a defence run that Twitter was not a ‘public electronic communications network’. However, this defence was dismissed by the lower court in Chambers v DPP and not re-addressed by the appeal court. Crown Court Judge Davies agreed that ‘the fact that [Twitter] is a private company is in our view irrelevant‘ and ‘the mechanism by which [the tweet] was sent was a public electronic network and within the statutory definition… Twitter as we all know is widely used by individuals and organisations to disseminate and receive information,, it is inconceivable that grossly offensive, indecent, obscene or menacing messages sent in this way would not be potentially unlawful‘ ([2012] EWHC 2157 (Admin) at [23]).
This is, with respect to the honourable judge, not an ideal interpretation. Defining Twitter and its ilk as ‘public telecommunications networks’ (PCNs) raises serious problems in terms of the bad fit with the EC Telecoms Framework and the complex web of laws and definitions associated with it and implemented into UK law. Watson and Ingram note, elsewhere on this site, that defining social networks as PCNs implies a great many duties on to them supervised by Ofcom including universal service, maintaining 999 emergency numbers, data retention for law enforcement agencies and so forth. In EU-speak, social networks are not categorised as PCNs but as providers of information society services – giving them a far less regulated status, and with benefits such as immunity from liability for third party content , subject to respecting notice and takedown.
But more than this, to define Twitter as a PCN is also to imply that the public authorities have a right to censor speech on Twitter more extensively than they are permitted generally to do so by Handyside and similar freedom of expression case law. This seems neither right nor appropriate.
This, in my view, is the nub of what has gone wrong with s 127 lately. Statute law designed:
(a) primarily to regulate one-to-one communications, rather than one to many broadcasting (whatever LJ Bingham said, it is clear almost all the wording of s 127 comes directly from statutes mainly intending to deal with malicious one-to-one phone calls or letters); and
(b) designed to safeguard a public utility built with public money,
is now being applied to a privately owned, though publicly accessed, many-to-many communications domain.
The DPP’s statement re projected guidance for s 127 shows the results of this tension. Section 127 is not drafted to fit modern guarantees of freedom of speech in public; and while both courts and prosecutors will no doubt do their best to interpret it in line with Article 10 jurisprudence, for them as well as defence lawyers and ordinary folk, this will be a hard and unpredictable task.
Here is a suggested way forward. Abolish s 127 with all its ambiguities and loose wording[14] and extend the Malicious Communications Act to apply to the whole of the UK[15]. That deals with one-to-one abusive electronic communications which have been acknowledged as a social ill since the days when all we had was the post and the telephone[16]. Rules criminalising abusive communications on a one to one basis do not chill public speech but do punish acts which create apprehension and invade privacy; the social ill is very different. But then stop, and have a decent debate about how to regulate one to many, public communications on the Internet and especially on social media. Such a debate needs to take into account excessive use of the PHA and the POA as well as s 127. And it needs to happen soon: the promised guidelines are a sticking plaster but as we all know, plasters do not stay on long.
In essence, what is needed is a manifesto for altering norms of behaviour on social media to reflect the demands of a civilised world for civilised discourse, while not prejudicing the unprecedented opportunity the Internet offers to expand social participation. This is a debate about emerging norms, about generational gaps, about digital natives, silver surfers, changing patterns of news acquisition and media literacy, as well as regulation – indeed it is a debate which, this writer suspects, may ultimately have almost nothing to do with law.
Lilian Edwards is Professor of E-Governance, University of Strathclyde. The piece first appeared in slightly different form on Professor Edwards’ blog Pangloss, http://blogscript.blogspot.co.uk/ .
[1] Parliamentary questions reveal 2,671 prosecutions under s 127 between 2003 and 2008 – see http://www.theyworkforyou.com/wrans/?id=2010-06-14c.1905.h . One assumes numbers may now be much higher given recent publicity.
[2] Chambers v DPP [2012] EWHC 2157 (QB).
[3] See http://www.guardian.co.uk/technology/2012/mar/21/man-racially-abused-collymore-twitter-spared-prison .
[4] See http://www.guardian.co.uk/technology/2012/sep/20/footballer-tom-daley-tweet .
[5] (1976) 1 EHRR 737.
[6] [2006] UKHL 40.
[7] See CPS Guidance at http://www.cps.gov.uk/legal/a_to_c/communications_offences/#an11 / and http://www.cps.gov.uk/legal/p_to_r/public_order_offences/ .
[8] See http://www.guardian.co.uk/technology/2012/mar/27/twitter-racism-taking-on-twacists . The most relevant section appears to be s 32 which involves racial aggravation of an offence under the Protection from Harassment Act 1997 (infra).
[9] Note though that the 1984 Act s 43 was replaced and abolished by s 127 of the 2003 Act.
[10] See LC Report No 147 (1988).
[11] See s 1(2A)(a) inserted by the Criminal Justice and Police Act 2001.
[12] Wikipedia cites Pipex as the UK’s first ISP operating dial up from 1991.See http://en.wikipedia.org/wiki/Internet_in_the_United_Kingdom .
[13] Walden also notes that this ‘does not appropriately describe our modern liberalised .. communications industry and therefore seems unsatisfactory’: I Walden Computer Crimes (OUP, 2007) at 3.199.
[14] A point not even canvassed here is whether s 127 should properly be used to control offensive images on the Internet given a variety of other legislation which already deals with such images. See Parliamentary response at http://www.theyworkforyou.com/wrans/?id=2008-04-22c.195543.h .
[15] It is instructive here to compare the recent Irish case of Matthews where s 13 of the Post Office (Amendment) Act, 1951 was used to prosecute a man for leaving ‘grossly offensive’ messages on the social network Bebo. The prosecution was originally successful but later quashed because the Act intended to apply to nuisance phone calls had been extended well beyond what the explicit words allowed , despite their having been amended in 2007. The case illustrates the dangers of confusing these two legal problems. See http://www.tjmcintyre.com/2008/07/bebo-bullying-and-law.html?m=1 .
[16] Indeed there is the beginning of a worrying trend to over extend legislation to deal with perceived gaps in the law regulating one-to-one transmissions as well. See recently R v GS [2012] EWCA Crim 398 so far only available in full at http://obscenitylawyer.blogspot.co.uk/2012/07/no-words-were-harmed-in-writing-this.html where the Court of Appeal found that a message sent one-to-one by IRC could constitute a breach of the Obscene Publications Act 1959.
