INTRODUCTION
Court judgments are public information but they also contain private data. Should they be kept in public or private hands, or is there a third way?
Most people seem to agree that public information should not only be publicly available, but also freely available. In the area of law, this is assumed to include not only legislation but also case law.
Not all court judgments can be said to amount to case law, in the sense of establishing new legal principles. But when they do, they become part of the law of the land, ignorance of which is considered no defence. It is probably impossible to know all of the legislation currently in force, and certainly impossible to know all the common law that has been developed, case by case, over hundreds of years. But we should at least be able to refer to it.Footnote 1
The question is how this information should be kept and made available for public access. Is public ownership the only way, can it be entrusted to private businesses, or can non-profit organisations offer an ideal solution?
WHAT IS PUBLIC INFORMATION, AND HOW DO WE FIND IT?
The case of legal information should be considered in the context of public information generally. This can include information created or provided by the government, the National Health Service, utility regulators et al, but also things like company registers and accounts, share prices, maps, business directories and so forth. In an even wider context, there is information which has been made public because it is in the public domain, such as news, artistic and intellectual content over which any private rights (such as patents and copyrights) have expired, broadcast content such as radio and TV, and content which has in some other way been deliberately placed in the public domain by its creators, such as the home page of ICLR's website.
Over this blooming meadow of public information, the general public like so many bees and butterflies may browse at will, gathering whatever nectar they want. But first they need to find the meadow, and then they need to find the flower.
In that distant-seeming epoch Before the Internet (or BI, if you will), you had to know where to go. It might be the public library, or Companies House, or a newspaper's business pages or, in the case of the Government, Her Majesty's Stationery Office. It might be your radio or TV. You might get it free, or might have to pay a fee.
After (post) the Internet (both AI and PI are taken, so think of your own epochal abbreviation) the situation has become a lot more fluid. If you happen to know where to look, all well and good: you can go to the relevant website directly, or follow a link from something related. If not, you can use a search engine to find it. For the vast majority of first-time queries, particularly if you are looking outside your area of expertise, the search engine is your first port of call, and in most cases will find what you want.
Search engines thus have a function that is to a large extent a public function, as an index or directory of public information. However, while the Internet itself may be neutral (there is some argument about this, but we needn't go into it here), search engines are not. They are businesses, and none more so than the pre-eminent search engine, Google.
Another kind of information that may be public, or subject to varying limits of distribution, is all the information on social media channels, such as Facebook and Twitter. Notwithstanding questions of privacy, it could be argued that these private businesses are also providing a form of public service.
THE VIEW OF THE COURTS
Two recent cases in the European Court of Justice have brought these issues into sharper focus. In the case of Schrems v Data Protection Commissioner, an Austrian law student is challenging the transfers of his personal data to the USA by Facebook, on the grounds that it would be subject to mass surveillance under US law, as revealed by Edward Snowden.Footnote 2
During the hearing (the court has yet to give judgment), the advocate for the European Commission made an astonishing suggestion: because the Safe Harbour agreementFootnote 3 signed between the EU and the US to protect the personal data of European users of US websites offers no real guarantee, any EU citizen not wishing their data to be passed to the American security services should close their Facebook account. This is almost like saying, if you don't want your conversations recorded, don't use the telephone.
The point is, yes, of course you could never use the telephone, and you could never use social media and you could never step into the public street, for fear of CCTV cameras, without a mask and hoodie (though this would draw attention to you more effectively than a business suit and spectacles). But the point is also that for anyone accustomed to using such services as part of their daily life, it is unthinkable to have to abandon them simply because mechanisms which are supposed to protect the rights of citizens are either not in place or, as in the Schrems case, assumed to be ineffective.
Telephone services, the management of CCTV supervision in public spaces, and social media such as Facebook are all privately owned and managed businesses, and we seem to be happy or at any rate enough of us democratically supportive of this situation to allow it to continue. In the context of personal data, however, there may be good reasons for thinking it is better protected in private hands than in public, given the threat of abuse (or undeclared misuse) comes as much from government-run security services, as it does from criminal hackers.
Now compare the Schrems case to that of Google Spain SL v Agencia Española de Protección de Datos (AEPD) (Case C-131/12) [2014] QB 1022, which established, inter alia, that a data subject has a “right to be forgotten” in certain circumstances, ie. a right to be delisted from search engine index results.
For the sake of argument, I'm going to assume that information published on the internet is (a) public information and (b) published there lawfully. In the Google Spain case, the information was a public notice which was required by law to be published in a newspaper or gazette. Included in the announcement was information relating to the credit status of the “data subject” (ie. the individual involved). It was an old story, and the purpose of the publication had been long exhausted by the time the newspaper's archive content came to be digitised and uploaded onto the internet. But in consequence of that public-spirited process, the content was crawled and cached by Google (and probably other search engines) and a snippet preview became visible in search results based on a search on the data subject's name. This information being long out of date, and no longer relevant to the data subject's current circumstances, caused him embarrassment. (It is like a more serious version of those stories a drunken uncle tells at his godson's wedding, dredged from the vault of long forgotten youthful indiscretion.)
But Google's initial response, and that of the free speech fundamentalists who supported the preservation of such information in a public space, objected to the exercise of any right to have it delisted on the grounds that it amounted to “censorship”.
An important aspect of the case is that Google is a private company, but is regarded by many as providing a public service. In this case, it would be no answer to suggest, as in the Schrems case, that you simply abstain from using Google. Who can abstain from using Google? In any case, unlike Facebook and the telephone, your abstention doesn't preclude your information appearing there, if it is in the public domain. The right to be delisted is the only way to abstain from appearing there yourself.
LAW REPORTS AS PUBLIC INFORMATION
One reason why the Google Spain case has a particular resonance in the legal sphere is the rehabilitation of offenders. I have personal experience of this – as a reporter, I hasten to add. I reported a case as a law report in The Independent newspaper in the early 1990s. It concerned a man who had pleaded guilty to an offence of dishonesty which, though comparatively minor, had resulted in a short prison sentence. However, it was one of those offences which, under the Rehabilitation of Offenders Act 1974, a person is after a certain period no longer required to mention or admit to in job applications (save for specific types of job). He had recently applied for a job without mentioning his conviction but his employer, searching for the man's name on the internet, had subsequently discovered my law report. In consequence he was not taken on.
It is not my job as a reporter to give legal advice, so I suggested he (a) consult an employment lawyer and/or (b) approach the Independent and ask them to redact his name from the archived report, which I think they could have done, if so minded, without asking my consent. If they'd asked, I'd have been happy to oblige. A law report (like a news report of legal proceedings) must be fair and accurate, not least to avoid any liability for defamation or contempt of court, but many cases are reported without disclosing names of litigants or defendants.
However, it is all a matter of degree. The case is also reported in the Weekly Law Reports, published by ICLR. These are not, however, susceptible to indexing by Google because, as a subscription service, they reside behind a paywall and in a discrete collection which Google does not crawl. The right to be delisted applies primarily to the sort of snippet preview that is most likely to come up in the index results of a search engine.
The story also illustrates one of the reasons why, when you search on Google for a case which is on BAILII, the British and Irish Legal Information Institute websiteFootnote 4, with which I imagine most readers of this publication will be very familiar, the Google hit list may show you a BAILII entry but it will not include any snippet of information. None is cached because, as the Google result will state, “A description for this result is not available because of this site's robots”. There is a good reason for this, which I will come to in a moment.
RAW LAW
The information on BAILII is public information. It consists mainly of post-2000 judgments of the senior courts in the UK and the European Courts of Justice and of Human Rights, but it also contains a collection of as-enacted UK legislation. This is all information in the public domain. It is, in fact, the sort of information the government could, and some think should, put online itself. In some cases it already does.
For example, all current legislation is available from Legislation.gov.ukFootnote 5, a site managed by the National Archives. Most of the old content on it was originally published in hard copy by the official printer, Her Majesty's Stationery Office (HMSO), which is now part of the National Archives. But new content is printed by The Stationery Office (TSO), part of the Williams Lea group, which in turn is part of Deutsche Post DHL (a commercial organisation). Unlike on BAILII, the legislation on Legislation.gov.uk is consolidated, ie includes amendments and repeals currently in force, although it is not completely up to date yet. (The intention had originally been to have it all up to date by the end of the 2010 Parliament, ie May 2015; but this has slipped and I believe the target date is now the end of 2015.)
In the case of judgments, while the content is generally believed to be covered by Crown copyright, the government does not publish them all, or even a fraction of the number appearing on BAILII (who in turn do not have all of them either). The Supreme Court and the Judicial Committee of the Privy Council each publish their own judgments and opinions on their own websites, as do the European Court of Human Rights and the European Court of Justice. But the Courts and Tribunals Judiciary websiteFootnote 6 only publishes a selection of judgments and sentencing remarks, chosen by reference to their likely media interest rather than their usefulness as precedents or for legal research.
That means that BAILII, which has all this and more (the only bit missing is the sentencing remarks, which have no precedential value) is the largest and best repository of free public information in the form of primary sources of law (raw law, if you will). This is law as it was made, with some links and additional information, but not turned into law reports or commentary. It is hugely popular, not just with journalists and academics but also with the judiciary and practitioners, who regard it as a first port of call for case law research, even if they have access to more comprehensive paid-for services. For the judiciary, its significance lies in its free public access. There is, in the Family Division and Court of Protection, practice guidance issued by the senior judiciary, directing that judgments be sent to BAILII for publication in pursuance of the transparency agenda initiated by the current President of those courts, Sir James Munby. Those directions could have specified publication on the Courts and Tribunals Judiciary website. But Sir James chose BAILII because, essentially, everyone knows and uses it.
In short, BAILII is as close as we can get to a public service, and yet it is not a public body. Nor, however, is it a private or commercial one. It is in fact a charity, managed by trustees.
THE POSITION OF ICLR
Also a charity, managed by trustees, is ICLR, the Incorporated Council of Law Reporting for England and WalesFootnote 7. This, too, performs a public function. It is a function it was set up to perform 150 years ago, in 1865, when the Inns of Court and the Law Society got together to promote a single, reliable body to publish law reports of the significant decisions of all the various courts of record then in existence, ie. courts whose decisions were capable of being binding precedents.
The courts covered included a number, such as the Court of Common Pleas, the Court of Exchequer and the Court of Crown Cases Reserved, which were swept away in the reorganisations of the 1870s. Moreover, the reports did not enjoy the monopoly the ICLR's founders may have intended, with rival series continuing, albeit in far smaller numbers, to provide their own coverage of some of the courts. What mattered, though, was that ICLR introduced systematic coverage. It applied known standards and criteria as to what was reported. And it provided the information at “modest” cost.
GUARDIANS AND GATEKEEPERS
Since ICLR and other publishers were not under a duty to publish reports of all the judgments of the court, there was no comprehensive public record of proceedings in court. For a case to be made public, it had to be published. That was an editorial decision. The same is true of factual or journalistic reporting, of course. But it meant that for many years ICLR performed a “gatekeeper” role in relation to legal decisions. If a decision wasn't reported by ICLR, the chances were that it wouldn't be known about and vanished into obscurity. Indeed, ICLR used to rather picturesquely describe its reporters as the “guardians of precedent”.
The advent of the internet has vastly reduced the ability of any one company or organisation to occupy this sort of gatekeeper role. But it has not removed it altogether. In the context of internet searches, at any rate in Europe, Google is the pre-eminent search engine and so occupies a gatekeeper role. Because its methodology is so dominant, if you can't find it on Google you are unlikely to find it anywhere else.
I'm not saying ICLR is the Google of law reporting, though that would be a nice position to be in! But its (former) gatekeeper role in relation to case law placed it in a similar position and imposed on it certain public responsibilities. To a large extent I believe it discharged those responsibilities, albeit not without missing the odd case that turned out to be of major importance, and not without some criticism, and in due course rivalry, from those who favoured better coverage in specialist areas.
In fact ICLR's approach going forward is going to be a lot more Google-like and inclusive: we plan to index all case law, even if we don't report it. The plan is to provide a single, comprehensive search facility for case law, wherever it is hosted. It may take a while to build up, but it will answer one of the central questions about public information: it may be available, it may even be free, but can you actually find it?
TRANSPARENCY, ACCESSIBILITY AND THE SEPARATION OF POWERS
Until recently, there were myriad government websites. The Government Digital Service (GDS) has over the past four years merged all government department websites and many other agencies and public bodies into one, called Gov.uk. There's no denying the ambition and scope of the project, but there have been questions about how easy it is to find content which used to be held on specific sites and is now in a great bran-tub of information. The risk is that the still separate Justice.co.uk website may go the same way.Footnote 8 The key problem is the difficulty of locating information within this vast Gov.uk site. Using its single field home page search box is like finding a specific size of brass screw in a 3-acre builders’ yard by asking the security guard in the carpark.
This well illustrates the difference between information being available, and it being accessible or transparent. Ironically, it is Google which comes to the rescue, because Google searches the whole site from outside, and its constantly updated algorithms (using your own browsing history) take you directly to the content you seek. This is both clever and a bit scary.
Currently, the Courts and Tribunals Judiciary has a separate site, as do the Supreme Court and Privy Council. In the case of the Supreme Court, under the deputy presidency of Lord Hope, it made a deliberate decision to go independent and separate itself from the “gov.uk” family. In 2013 it changed its web address to www.supremecourt.uk from the old www.supremecourt.gov.uk which it had initially been forced to use, according to Joshua Rozenberg in The Guardian, “implying that it was part of the government rather than the final safeguard against unlawful government behaviour”.Footnote 9
This is a point about the separation of powers which could equally well be made in relation to all judgments. In other words, while the Courts and Tribunals Judiciary website (www.judiciary.gov.uk) is not yet part of the Gov.uk umbrella, there may be a good argument for saying it should not even include “gov.uk” in its URL – a signal to reinforce the independence of the judiciary. These days, an organisation's URL is a key part of its branding and image.
While it remains part of the government informational fold, however, support for an independent repository of judgments (many of which may, as Rozenberg suggests, be critical of the government and public bodies) will continue to be strong, and BAILII will continue in its role as the trusted, trustee-led, not-for-profit custodian.
One of the reasons BAILII doesn't allow Google to snippetise its content is the risk of inadvertent disclosure of content that should be redacted. BAILII is the go-to site for family law, not least because of the President of the Family Division's preference for it and directions to use it. But many of the decisions involve children and vulnerable parties, whose identities should usually not be disclosed. In most cases this is not a problem, as the judgment will be issued with a tantalising title such in “In the matter of G (Children – Habitual residence)”. But there is always a risk that a child may be identified accidentally, or by what is known as “jigsaw identification”, piecing together unusual facts that may have remained unredacted in the judgment. BAILII is very quick to remove or update content as soon as notified by the court or by parties or their legal advisers, usually erring on the side of caution. But a cached version of the judgment on Google would not reflect that change of situation for some time. Hence the protective robots.
Incidentally, problems with jigsaw or even outright contemptuous (of the court) identification of children in family cases does go on all the time, on Facebook and various online forums. Disappointed litigants are usually the ones responsible for this, but in many cases they don't understand why they can't talk about their own children in the same way all their friends post pictures of birthday parties and barbecues. Except in the most egregious (even nutty) cases, it is tricky to police and trickier still to enforce.
OTHER JURISDICTIONS
An exhaustive comparison of the position in other jurisdictions is outside the scope of this piece, which is already overlong. But it is worth looking at a couple of examples which the UK might consider following.
First, there is the new US Supreme Court website launched in 2010. As with the UK Supreme Court, the site is run by the court itself, separate from the government. What I like about it is the sheer volume of information about past and current cases. You can search by the “docket” number of the case, the case title, or other words or numbers included on a docket report of any case going back to 2001. The report lists all the filings and their dates. You can read slip opinions (first drafts of judgments) and court orders. You can also read merits briefs on a linked website run by the American Bar Association, which is similar to being able to read skeleton arguments (though these are much fuller). And you can listen to an aural recording of the oral hearing. These are also available as a transcript.
For anyone reporting a case in the Supreme Court, this offers an embarrassment of riches. In similar vein is the Canadian Supreme Court. This offers a “statement of policy” about its approach to transparency which seems a good model:Footnote 10
The SCC will provide access to court records in any judicial proceeding before the Court to members of the public, media and parties to the judicial proceedings in a manner that balances the constitutional requirement of open courts against other rights and interests of the public and participants to judicial proceedings, namely privacy and security of individuals and the proper administration of justice.
Further examples of how it could be done are offered by the well-funded European Courts of Justice (in Luxembourg) and of Human Rights (in Strasbourg). In the case of the latter, the available material related to a case includes legal summaries and press releases as well as all the judgments. In the case of the former, there is both a general website for EU laws www.eur-lex.europa.eu (which offers legislation and the official journal) and a specific site for case law, www.curia.europa.eu. Because both sites are multi-lingual, there is an element of duplication in much of the collection, and searching can be laborious, but there is no doubting the comprehensiveness of what's on offer.
Our own Supreme Court is moving in the right direction but in halting steps. Hearings are recorded and may be viewed up to a year laterFootnote 11, and judgments are also aired on its YouTube channelFootnote 12 but the archive of legacy recordings more than a year old will only be accessible, via the National Archives, after a five-year wait, which seems baffling.Footnote 13 The website does provide some details of current cases, including links to the judgment below, but no access to the written arguments or pleadings. Other UK courts lag far behind, with barely more than you can read on a daily cause list.
A VISION OF THE FUTURE
Before leaving this subject, I want to sketch out a vision of how, ideally, the courts should approach the online realisation of open justice. I would suggest they provide, on their website, access to four types of information relating to litigation (whether cases are heard in a physical court or, as recently envisaged, by some form of online dispute resolution).
(1) The process documents. These would include the claim form, pleadings, appellant's notice, skeleton arguments, affidavits, and perhaps even some exhibits. In other words, what you might include in the “core bundle”.
(2) The hearing, accessible by way of audio or even video recordings, together with a transcript (mechanically produced, perhaps subject to correction by user-annotation, like a wiki).
(3) The judgment, initially in draft form as handed down, subsequently the approved version.
(4) The orders of the court, especially including awards of damages, costs and any injunctions granted. (The nearest we have to this, at present, seems to be county court judgments against local businesses, which may be relevant to their solvency.)
Obviously anything that should be redacted, up to and including the entire case, could be shielded from public view. But such a system would offer benefits not just for legal researchers but also the court staff themselves (for which much of this is already, I assume, available internally).
The logical extension of such a facility would be for the judgments to be linked (rather in the manner the US Supreme Court links to the American Bar Association for access to merits briefs) to the official Law Reports of those cases which justified the added value treatment. But there's no reason why ICLR, as the embodiment of the original vision of reliable public access to case law in 1865, should not play an even more integrated, perhaps even pivotal, role in such a project.
CONCLUSION
Our assumptions about how public services should be provided have changed in the last 30 years, from a model of public ownership to one of public regulation of competitive private provision. This may work well for telecoms and energy utilities, less well for public transport, and tends to meet with widespread ideological objection when it comes to education, letters sent by post and, perhaps most vociferously, health services. In the case of information, however, the key driver seems to be accessibility.
There's also a divergence based on what you call it. When you talk about facts as information, everyone thinks it should be free and easy to find and use. When you talk about facts as data, especially personal data, then it becomes something to worry about. We're happy to put all sorts of rubbish onto Twitter or Facebook, without a second glance, but when someone tells us the same information is being sedulously stored and mined by the NSA or GCHQ we start to complain about invasions of our privacy.
There's a balance to be struck, obviously. And this tension between what is public and what should remain private, which is at the core of the Schrems and Google Spain cases (among many others), highlights one of the risks associated with the custodianship of public information. Is it safe?
In this context, the word “safe” has more than one meaning. Personal data should be safe from prying eyes. But public information should be safe from interference, suppression, censorship etc.
Where that public information is the law of the land, the essential requirement is that it should be accessible. For legislation, which is the expression of the legislature, all that is required is that it should be published in some way. In the past, it was sold at a modest price in hard copy by HMSO (and now by privately-owned TSO). Now, it is available online, in a form which is approaching full consolidation even if it hasn't quite got there, from a free website managed at public expense by a public body, the National Archives. It is also available, as enacted, from BAILII, which is free, but is run as a not-for-profit trust.
For case law, the position is more complicated. The custodianship of the information includes a need to protect against potential harm to the interests of other parties. The primary duty to protect those interests must fall on the court, and in an ideal world it would also be the courts that would control and publish the information, including all their judgments and orders, but also (ideally) the preparatory paperwork and the hearings. This is what would best satisfy the “open courts” principle cited by the Canadian Supreme Court website.
To the extent that courts do not perform that function, it can be performed by others. In the past, the ICLR sold at a modest price reports of those decisions of the senior courts that were considered worthy of preservation, out of the general run of legal business, because of their value to students and professionals of the law. In the present day, ICLR continues with that project, but other commercial publishers have added their efforts, widely, to the store of content; and BAILII now offers a more comprehensive collection, though without the added value of full text law reports.
Whether the judgments are published by the courts, the law reporters or BAILII, the risk of harm to affected parties is minimised, because these organisations are conscious of the need to protect the interests of data subjects or vulnerable parties. What they also need to be is independent, and that is why the better model, if the courts do not publish judgments themselves, is for those who do so to be unaffected either by government policy or commercial considerations. That suggests the current model of a trustee-governed not-for-profit organisation like BAILII or ICLR is as good as we can have, for the moment at least, in terms of ensuring both (a) respect for privacy and (b) accessibility at no or low cost.
