Forget me, not: the extent of the right to be forgotten

Make an embarrassing revelation online that you later come to regret and, at present, you have no way of ensuring its deletion from the all-remembering internet. This could change following the announcement by EU justice commissioner Viviane Reding in March that the 'right to be forgotten' would soon be enshrined in data protection law.

Under the new law, compromising or simply unwanted information will be 'forgotten'. Does the principle behind digital amnesia extend not only to Facebook indiscretions but also to minor criminal offences? The case of R (T) v Chief Constable of Greater Manchester Police [2012] EWHC 147 suggests that those with criminal records might have the hope that their trespasses will truly be forgiven by being forgotten.

The phrase 'right to be forgotten' originates in civil law countries, in particular in French law's droit à l'oubli, a convict's right after rehabilitation to have his record expunged (see Rosen, 2012, 64 Stanford L. Rev. Online 88). English law contains a similar, though less powerful, amnesty: section 4 of the Rehabilitation of Offenders Act 1974 (ROA) allows those with spent convictions to claim lawfully that they are of good character, with a spent conviction not being a 'proper ground for dismissing or excluding a person' from employment. However, under sections 113A and 113B of the Police Act 1997 (PA), all criminal records certificates issued by the Criminal Records Bureau must contain both spent and unspent convictions and cautions; and, for many forms of employment, the protections guaranteed under section 4 ROA 1974 are removed by the Rehabilitation of Offenders Act 1974 (Exceptions) Order 1975. Hundreds of thousands of people applying and volunteering for a range of positions from doctors and teachers to childminders must for the rest of their lives either disclose or have disclosed all of their wrongdoings, however minor and ancient.

Should one's misdeeds be forever remembered? The claimant in T argued before Parker J in the Administrative Court that they should not. T, a 20-year-old sports studies student, was, as part of his course, required to obtain a criminal records certificate. The certificate contained a spent caution '“ a warning administered to him by police after he admitted, at the age of 11, to stealing two bicycles. T contended that the PA 1997 was not compliant with article 8's right to privacy as it allowed no review of whether convictions and cautions should be included on criminal records certificates. He also argued that the 1975 order exempting those applying for specified positions from the protections ordinarily afforded to rehabilitated offenders was ultra vires and violated article 8.

System with no exceptions

Parker J was greatly troubled that 'a system that allows no exceptions imposes a very heavy cost in terms of effect on the fundamental rights [to privacy]'. The police had themselves recognised the injustice of life-long disclosure of criminal records. The Association of Chief Police Officers' 2006 Retention Guidelines for Nominal Records on the Police National Computer formerly regulated the 'stepping-down' procedure whereby convictions and cautions could be made non-disclosable outside the police force. Sadly, this procedure was suspended after Chief Constable of Humberside v Information Commissioner [2009] EWCA Civ 1079, where it was held that the police were obliged to disclose all records of convictions and cautions to the CRB without exception.

Despite his discomfort at the blanket disclosure of records, Parker J dismissed T's claim. He ruefully felt constrained by his reading of the Supreme Court's decision in R (L) v Commissioner of Police for the Metropolis [2009] UKSC 3 to conclude that the present system was article 8 compliant. L concerned not convictions and cautions appearing on certificates but 'non-conviction material' (arrest records and police intelligence). Parker J inferred that the majority in L had deemed the disclosure of spent convictions under the PA 1997 compatible with article 8. The judge made it clear that, were it not for L, he would have declared the regime incompatible with article 8 because of the absence of a review mechanism.

That dealt with the state's direct role in the disclosure of previous convictions. Parker J sympathised less with the claimant's second contention in T that the 1975 order was ultra vires and violated article 8 because it required rehabilitated offenders themselves to disclose spent convictions and cautions and allowed a spent conviction to be a 'proper ground' for excluding a person from employment.

Parker J denied that the state was obliged to shield ex-offenders from having to disclose spent convictions when applying for positions of trust. Applying Mosley v United Kingdom [2011] 53 EHRR 30, he reasoned that a novel positive duty to uphold privacy should only be imputed where there was, at least, an 'important facet of personal identity' at stake. His refusal to recognise a positive duty to protect ex-offenders jarred with his approval of the words of Lord Hope in L: 'As [a conviction] recedes into the past, it becomes a part of the person's private life which must be respected.' How can a long-spent conviction not be an 'important facet of personal identity'?

What will the Court of Appeal make of T? Will it share Parker J's reading of L? And even if not constrained to conclude that the current system is article 8 compliant, will the Court of Appeal endorse the status quo?

The domestic judiciary has been wrong about privacy and police archives before: the House of Lords unanimously supported the blanket retention of DNA only for the European Court of Human Rights to damn it unanimously (R (Marper) v Chief Constable of South Yorkshire [2004] UKHL 39; Marper v UK (30566/04)). Like Marper, an indiscriminate encroachment on privacy appears to be as central to T as it was in R (F) v Justice Secretary [2011] UKSC 17, where the Supreme Court held that the lack of any review of a life-long requirement for sex offenders to notify police of their whereabouts violated article 8.

Is Westminster heading for another conflict with the domestic judiciary or the Strasbourg court over police databases? Probably. Despite the forewarning offered by T, the coalition government's Protection of Freedoms Bill is trundling through parliament without any attempt to correct the privacy deficit in the disclosure of criminal records.

The Protection of Freedoms Bill includes some modest changes to the criminal records system. Criminal records certificates will be sent to prospective employees before prospective employers to ensure that contested material can be challenged before any damage is done. The test for inclusion of police intelligence in enhanced criminal records certificates is to be tightened along with a right to a review of the inclusion of non-conviction material. There is likely to be a reduction in the number of CRB checks issued after the scope of the vetting for those working and volunteering with vulnerable people is narrowed. The secretary of state will also be able to order the 'deletion' of convictions and cautions for consensual homosexual acts historically seen as offences. The bill's definition of 'deletion' departs somewhat from the word's literal meaning: the record will remain on the Police National Computer but with a note appended that the conviction should be disregarded. The conviction will then become non-disclosable. Could convictions and cautions such as T's be similarly 'deleted'?

No consensus

What is left undone by the bill? Section 56(5) of the Data Protection Act 1998 has not been brought into force which would make it an offence for unscrupulous employers to force prospective employees to apply for CRB checks for positions fully protected by the ROA 1974. And, although not a 'proper reason' to refuse employment, where is the remedy for job applicants rejected on the basis of spent convictions? And, of course, where is the review mechanism for the disclosure of convictions and cautions ?as in T?

How would a review system for convictions operate? In Germany, convictions are recorded on the Bundeszentralregister for between five and 20 years depending on the sentence's severity. Could offences in England also have expiry dates? Operational questions aside, the need to reconsider the management of criminal records has become more pressing because of the implementation last month of the European Council Framework Decision on Exchange of Information from Criminal Records (2009/315/JHA), which will, through the European Criminal Records Information System (ECRIS), ease the transmission of criminal records between member states for the purposes of sentencing. As Parker J pointed out in T, there is no consensus within Europe regarding criminal records. Will the ECRIS system only go to complicate an already complex system? And Does the deletion of criminal records in Portugal, France, Greece and Germany unfairly prejudice Britons whose state refuses to delete their records?

The absence of a European consensus on convictions was illustrated by a Grand Chamber decision in February this year, Axel Springer v Germany (App No 39954/08). On the surface, Axel Springer was a healthy assertion by Strasbourg of press freedom. The German courts' injunction against Axel Springer publishing the conviction of a well-known television actor for the possession of drugs was overturned in Strasbourg: the actor's privacy had to give way to freedom of expression. What Axel Springer reveals is that Germany's courts prized an actor's privacy over the publication of the fact of his recent conviction. German judges would have had in mind the 1973 Lebach case (35 BVerfGE 202), in which Germany's highest court held that the getaway driver in a robbery of a German army depot by radicals in the 1960s had the right, as part of his resettlement in civil society, not to be named in a television show. An English court would be unlikely to share this approach.

Axel Springer also highlights the problem of someone simply telling others of your spent conviction. Under section 8 of the ROA 1974, it is defamatory to publish another's spent conviction but the defence of justification will apply unless malice is shown. And what of press archives? When the ROA 1974 was drafted, ferreting out a press report of another's conviction archived on microfiche would have been a prohibitively difficult task. Now, type ?an ex-offender's name in Google and the search engine will remember instantaneously local newspapers' reports of their crime, especially if their crime and name were unusual.

Preventing the proclamation of your spent convictions rests on strained interpretations of the ROA 1974, the common law, post-Campbell and post-Mosley jurisprudence on confidentiality and privacy and the Data Protection Act 1998. As Eady J noted in KJO v XIM [2011] EWHC 1768 when refusing summary judgment to a nephew attempting to prevent his uncle from informing potential employers of his spent conviction for forgery, 'the boundaries in [this] area of law are unclear and developing'. And should there be a distinction between forgiving by forgetting, as in T, and rewriting history as in Axel Springer?

T could be the beginning of a long reconsideration of when to forgive and when to forget.