The General Secretary of the Labour Party, Jennie Formby, sent a letter to its Deputy Leader, Tom Watson, this week, expressing concern about Watson’s request that complaints about anti-semitism be copied or forwarded to him so that he could ensure that they were being dealt with properly. Formby outlined concerns about the effect on the complaints process that I am not qualified to answer, but she also raised the spectre of GDPR and Data Protection, and here, I am somewhat sceptical of her arguments.

Formby is right to say that political opinions are classed by the GDPR as special categories data and so require extra protection as compared to ordinary data like a name and address. I’m surprised that she didn’t also mention that many of the complaints would also include the religious or philosophical beliefs of complainants or the complained- about, as well as possibly their racial or ethnic origin. Why Formby didn’t want to highlight the religious and racial dimensions of complaints about anti-semitism is a bit of a poser.

To claim that Watson isn’t taking a risk by soliciting this data when he wasn’t already is clearly false. The best way to avoid Data Protection problems is not to process data in the first place, and as the third GDPR principle requires data minimisation, the safest choice for Watson is to trust the process and not receive any data. The problem arises if he doesn’t, or if he feels responsible for ensuring that it is working. If you think he’s operating purely politically, that’s your choice but I’m going to give him the benefit of the doubt. I’ve already had several heated disagreements on Twitter about this, mainly with people who are certain that he’s going to breach GDPR but uncertain about which particular element is in play.

So here’s my opinion. Tom Watson, in his capacity as Deputy Leader of the Labour Party, can process special categories data concerning complaints about anti-semitism, and Jennie Formby is wrong to argue that he cannot. It’s entirely possible that Mr Watson will follow the GDPR principles to the letter, and it’s entirely possible he will make an almighty cock-up of it. If he does, he should face the consequences. The Labour Party is no stranger to dodgy data dealings – it bought data that had been unlawfully obtained on its behalf by Emma’s Diary for the 2017 General Election, and while Emma’s Diary got fined, Labour didn’t. Sometimes, DP gets breached and nothing happens.

Formby said the following:

“The suggestion that you as an individual data controller should receive and store data relating to complaints unrelated to your personal role as an MP, on a private email address, or indeed any other system, is completely unacceptable and exposes you, and the Party, to significant compliance risks.”

Like all MPs, Watson is an individual data controller, but only when acting as MP for West Bromwich East and the constituency issues associated with that role. If Watson was acting as an MP and party members chose to forward their own complaints to him, or provide complaints made to them by others, it would be odd, but the Labour Party would not be responsible as Formby claims. Watson would be the controller. However, Watson is the party’s Deputy Leader and it is plain that he was acting in that capacity when he sought to receive the complaints – this is plainly a Labour Party matter, not an issue concerning the fine folk of West Brom.

If something in Labour’s constitution explicitly forbids the Leader or Deputy Leader from having direct involvement in, or oversight of, complaints, Watson has a problem. Formby’s letter clearly sets out her opposition to political involvement in the complaints process, and if she can back that up with a clear reference to the party’s formal rules, any argument that I might make in Watson’s favour is severely weakened. Labour’s formal internal rules have great significance for whether his processing is lawful. But if there isn’t, as Deputy Leader, I can’t see how Watson’s claim to determine the purposes for which the party uses data isn’t valid. In all big organisations, senior people can decide how and why data is used. The controller isn’t one person, it’s the organisation itself, and so logically, more than one person is involved in determining the purposes. If the Deputy Leader isn’t an appropriate person to make these decision, who is? Is it just Mr Corbyn?

There are at least two people’s personal data involved in any complaint Watson wants to see – the complainant and the subject of the complaint. Watson has to justify the processing of the data generally, and if the data is special categories, he has to find an exemption that allows him to process the special data.

If any person chooses to forward or copy their own complaint to Watson or his staff, they plainly consent to him processing their data. He can process any data about their political, religious or philosophical beliefs or ethnicity on the basis of their explicit consent. Watson might struggle to demonstrate he has explicit consent unless the email says ‘I explicitly consent for you to process data about my religion, politics or ethnicity’, but in the real world, it’s impossible to believe that the Information Commissioner or the courts would uphold a complaint from the very same person who forwarded their own complaint. It’s nonsense.

If the complaint is submitted to Watson by a third party, this is more tricky, unless of course the person forwarding or copying the complaint (presumably an MP or other party member) gets consent from the complainant. If that consent exists, Watson is in the clear. If not, he must establish a lawful basis to process the data. In my opinion, he has a legitimate interest in receiving and monitoring complaints about anti-semitism in a party of which he is Deputy Leader, especially when a Jewish Labour MP has just left that party because of anti-semitism. Watson would need to evidence the legitimate interests assessment, but I believe he could make it out. It is surely the role of a Deputy Leader to want to make sure that complaints are being dealt with properly, especially when the issue is as important or potentially damaging as this?

If he doesn’t have consent from a complainant and the complaint contains special categories data, Watson has another hurdle to clear in terms of a special categories exemption. However, the GDPR allows the processing of special categories data in the following circumstances:

“processing is carried out in the course of its legitimate activities with appropriate safeguards by a foundation, association or any other not-for-profit body with a political, philosophical, religious or trade union aim and on condition that the processing relates solely to the members or to former members of the body or to persons who have regular contact with it in connection with its purposes and that the personal data are not disclosed outside that body without the consent of the data subjects”

As long as a complaint was made by a current or former member, Watson can argue that his monitoring of the process is carried out in the course of the party’s legitimate activities. Alternatively, he could argue that it is necessary in the substantial public interest. I believe that ensuring that Labour is taking anti-semitism seriously meets that definition, although Watson also needs legal authorisation from the Data Protection Act 2018 to rely on that exemption. The authorisations include ensuring that equality of treatment is maintained (including the treatment of people with specific religious beliefs). I don’t think Watson needs this, but it’s there.

As far as the data of the complained-about goes, Watson plainly doesn’t have their consent and has no hope of obtaining it. However, I believe again that he has a legitimate interest in gathering the data, and if the data includes special categories like political opinions or religious beliefs, he can rely on the legitimate activities exemption quoted above. Once again, if you disagree, it’s worth noting that another exemption from the prohibition on processing special categories is the fact that the person clearly made the data public themselves. Many of the complaints about anti-Semitism come from comments made online or in public meetings and speeches – the meat of the complaint is very likely to concern public statements or utterances from the complained-about. You cannot object to Tom Watson reading a complaint about your tweets because you tweeted them into the public domain.

If Watson goes ahead with this, he has plenty of work to do. I do not believe that any of the above removes the obligations to comply with GDPR’s transparency requirements – everyone whose data he receives needs to be informed about the fact that their data will form part of the complaints process and Watson’s review. The data must be secure, not used for anything other than the complaints process, and of course, anyone whose data is being processed has rights over their data. Given the less than congenial state of intra-Labour relations, it’s not hard to imagine that the Deputy Leader might be left to his own devices should a nasty subject access request come knocking.

Of course, all of this is bollocks. The Labour General Secretary doesn’t really care about Data Protection (neither did her predecessor). This is politics. Anyone who describes their processing arrangements as ‘elaborate’ is making it up as they go along. No political party has a good record on Data Protection, which is why it’s a shame that the Information Commissioner is so shy about taking them on. Even though I believe that Watson wants to root out anti-semitism in Labour’s ranks, I wouldn’t be surprised if part of his aim is to send a message to Formby about the process.

There is, however, a solution to the whole mess, should anyone wish to take it up. If Watson withdraws his request to see all the complaints, Formby could offer to supply him with pseudonymised versions of them. That way, Watson could carry out an appropriate supervisory role, ensuring that this most vital of tasks is carried out efficaciously, but at the same time, no directly identifiable personal data would be made available to him, and he could assure Formby that he would not attempt to identify the parties. The risk would be low, the data would be protected, and surely nobody could object to an elected Deputy Leader keeping a watchful eye on an issue that is so very important?

SPECIAL RULES FOR COMMENTING ON THIS POST

1. Any comment containing the word ‘Israel’ goes in the bin.
2. Any comment that is about who funds who goes in the bin.

Thank you, and good night.

A couple of days ago, the training company IT Governance reported that the Information Commissioner’s Office had banned Keith Hancock, director of a Manchester lead generation company, from being a company director for four years. The ICO had previously fined the company (Lad Media), and this was the follow-up. All good stuff, you might say, perhaps even a riposte to those awful people who say that the ICO never does anything. Except it isn’t true. The ICO didn’t ban anyone because they don’t have the power to do so. The action was taken by the Insolvency Service with the ICO’s assistance. Weirdly, the IT Governance’s scribe used quotes from the Insolvency Service’s press release without either reading or understanding what it said.

UPDATE: demonstrating the lack of class that is ITG’s hallmark, the story has now been updated without any reference to the fact that it had been wrong, or that they needed me to correct them. This is what it used to look like:

I don’t expect IT Governance to get things right (their sales director once claimed that there had been GDPR fines of 6.2 billion against Facebook and Google), but you’d hope for higher standards from, say, the chairs of four Parliamentary Committees, right? Right? A week or so ago, a distinguished group of Parliamentarians (and Damian Collins) wrote to Jeremy Wright, Secretary of State for Culture, Media and Sport as part of a campaign to change the way the ICO is funded. The idea is that the ICO would get to recover the costs of its investigations from those found to be in breach of Data Protection law, and has been promoted by the Durham-based marketer Russell James. I think it’s a bad idea – it would require the ICO to record and cost the time they spend on every investigation, it could dissuade organisations from appealing ICO decisions (which is bad for everyone as ICO decisions need to be tested), and even where it was applied, it would see the ICO bogged down in arguments about how much they actually spent.

Leaving that aside, the letter itself is amateurish and inept. Several times, it refers to organisations being “found guilty“, something which only happens in criminal cases, thus ignoring the fact that much of the ICO’s work carried out under civil not criminal law. In similar vein, it refers to “data crimes“, a phrase presumably culled from Liz Denham’s misleading soundbite “data crimes are real crimes” (they’re not). This means that the scope of the letter isn’t clear – are they referring to civil breaches (which aren’t crimes), or are they referring to criminal offences, which in the ICO’s world are usually committed by individuals rather than organisations? I find it hard to believe that Dominic Grieve and Yvette Cooper would sign a letter than hadn’t been properly thought out, but as it turns out, they signed a letter that hadn’t even been proof-read. The penultimate paragraph includes a sentence that plainly has words missing “To strengthen the enforcement mechanism, and thus provide maximum credibility to the ICO should be able to recoup the costs of investigations…“, and most damning of all, it opens by describing the ICO as the ‘Independent Commissioner of Information’, which as Neil Bhatia pointed out would be make them the ICI, not the ICO.

UPDATE: a commenter below argues that I should not describe them as ‘civil’ breaches; rather, they should be described as breaches of administrative law. Technically, I think this is correct, although the point I was making is that they are definitely not crimes. I have made the entirely avoidable mistake of listening to the Information Commissioner, who describes them as ‘civil monetary penalties’, e.g. here. I will endeavour not to make the mistake of listening to the ICO again.

Here we have senior Parliamentarians putting their name to a letter that is badly written and incoherent, asking for changes to the funding of a regulator they can’t even accurately name. Russell James told me that the letter was drafted by Tom Tugendhat’s office, but it’s plain that nobody involved in its creation knows anything about Data Protection.

Bullshit is everywhere. In the same week as the ICI letter, Privacy International published a piece responding to Will.I.Am’s well-intentioned but counter-productive ideas about monetising personal data to benefit individuals. The piece included several completely false statements, including that fact that Cambridge Analytica had been fined by the ICO, and that Professor David Carroll had successfully sued the company to recover his data. I took this up with them and they attempted to correct the piece, but in doing so, they made it worse. The correction says “A previous version of the piece implied that Cambridge Analytica has been fined for their involvement in this scandal. The piece was updated on 7.02.2019 to make the text less ambiguous.” The problem with this is that the previous version didn’t imply anything: it said explicitly that Cambridge Analytica had been fined, and they haven’t. The correction goes on to say “The company has been fined for failing to respond to an access request by the Information Commissioner’s Office (ICO)”. It hasn’t. The ICO has prosecuted SCL Elections (not Cambridge Analytica) for failure to comply with an enforcement notice. Despite that famous raid, ICO hasn’t fined Cambridge Analytica or SCL, and the chances that they ever will be are roughly equivalent to me being invited to tea with the Commissioner.

You could be forgiven for asking ‘does it matter’? Does it matter that people get things wrong as long as their heart is the right place? Russell James told me repeatedly that it didn’t matter that the MPs’ letter was full of errors; what matters is that the letter was sent and the wheels are turning. It’s true that pedantry and point-scoring are an unhelpful feature of Data Protection discourse. However, there’s a difference between a conversation and a formal letter or article. More importantly, there’s a difference between pedantry and precision. If you’re talking about privacy impact assessments in the context of the GDPR and I correct you to say it’s a Data Protection Impact Assessment, I’m being a dick. We both know what you mean, and my correction adds nothing. If everyone thinks that the ICO fined Cambridge Analytica when they didn’t, it stops people asking questions about why Wilmslow has spent £2.5 million on an investigation that has resulted in a dodgy fine against Facebook and some mediocre PECR penalties on Arron Banks’ ramshackle empire. If MPs don’t understand the laws that they’re signing letters about, how do we know that they’ve scrutinised the campaign that they’re backing?

The problem is, the Commissioner’s Office are as bad as everyone else and sometimes they’re the source of the infection. Last week, the ICO tweeted that they’d fined Magnacrest Housing, when in fact, it was a court that issued the fine. When SCL Elections pleaded guilty to failing to respond to the ICO’s Enforcement Notice, the Commissioner proudly announced that they had taken action against Cambridge Analytica – although admittedly part of the same group, they’re two different companies, and nobody at the ICO wants to be precise about that because Headlines. The Commissioner herself has repeated the ‘data crimes are real crimes’ claim on many occasions, despite the fact that it’s both misleading and an unhelpful over-simplification. Denham endorsed a book she hadn’t read as “authoritative“, describing its author as someone who “flew the flag for data protection many years before it broke into the mainstream with the GDPR” when he was in fact a PR guy who jumped on the bandwagon.

Denham doesn’t even seem to be overly precise about what her job is – she was quoted by her corporate Twitter account yesterday as saying “What’s technically and legally possible is not necessarily morally sustainable in our society. That’s what the debate is about.” Denham is a regulator – it is her job to enforce the law. As several people have told me since I complained about the statement, Data Protection is principles-based and therefore not as fixed and binary as other areas of the law. I cannot deny this, but even taking it into account, the slippery and complex aspects of DP are still ultimately in the “legally possible” part of the Venn diagram. It’s none of the ICO’s business whether companies do things that are legally possible but morally questionable. If a company breaches DP or PECR, the ICO should take action. Either Cambridge Analytica broke DP law in the UK and the ICO can prove it, or they didn’t. It doesn’t matter that Alexander Nix is a smug gobshite because being a smug gobshite is not a breach of DP law.

We live in an era of fake news where the President of the United States routinely gaslights the world and AI can write prose like a human. The truth matters. Facts matter. Accuracy matters (it’s one of the GDPR principles after all). We all make mistakes. I do it all the time, and the best I can do is hold my hands up and do better next time. But when you’re a big organisation with a much bigger audience than some show-off trainer like me, when you’re an MP asking for a change in how a regulator is run, and especially when you’re charged with regulating something as important as the protection of personal data of 60 odd million people, it matters a lot more. You have to care about the facts because so many people are listening, and you have to take the time to get it right.

And now, in the time-honoured tradition of this blog, I will hit ‘Publish’ and spend the next hour spotting all the typos I’ve made and editing them out before anyone notices.

With the sensitivity for which they are rightly renowned, the Home Office chose to celebrate Christmas by tweeting a cheery video full of beaming millennials, promoting the new ‘settled status’ registration scheme for EU nationals who want to stay in the UK after Brexit. People who have made their home in the UK have to register and pay for the privilege. Setting aside the crass, thoughtless way in which the scheme was promoted, concerns have been expressed on social media about the Data Protection implications, especially as regards how data is used and whether it complies with GDPR and the DPA 2018. There is an interesting sentence in the documentation: “we may also share your information with other public and private organisations in the UK and overseas“. The people behind the @the3million twitter account made an FOI request about this, and the Home Office have refused to confirm the identity of the organisations in question. They relied on S31 of the FOI Act, which allows information to be withheld if (among other things) disclosure would or would be likely to prejudice “the operation of the immigration controls“.

S31 requires the Home Office to demonstrate a causal link between disclosure and prejudice, and has a public interest test that allows for disclosure if the public interest in doing so outweighs the public interest in withholding. So while the Home Office picked the right exemption, their decision to refuse could be challenged. The ICO doesn’t have a strong record of overturning these kinds of decisions, so the fate of any complaint is hard to predict.

But what’s that? Surely individuals subject to this process have GDPR rights, and can find this out for themselves via a subject access request? Two elements of GDPR would appear to assist – Article 13 requires the Home Office to specify “the recipients or categories of recipients” to which personal data will be disclosed in order to be transparent, while Article 15 gives the subject a right to the same information on request as part of a subject access request.

Except they don’t. I’m certain that the wording I have seen doesn’t comply with Article 13 because even the ‘categories’ bit would only work if it was clear what types of recipients are involved, and it’s plainly not. However, the GDPR allows for exemptions, and there is an exemption that the Home Office managed to get through Parliament in the DPA 2018 which allows them to keep the identity of the recipients secret. Schedule 2, Pt 1, (4) says that both transparency and subject access rights can be set aside if applying them would or would be likely to “undermine the maintenance of effective immigration controls“. If the Home Office don’t want to tell people going through the process who their data will be shared with, this exemption allows to do so. They have to believe that transparency will undermine effective immigration control, but this is the Home Office – they probably do believe that.

So what recourse do EU citizens have? They could, of course, challenge the Home Office approach by either taking them to court or complaining to the Information Commissioner. The Commissioner could decide that the application of the exemption was incorrect (as they could with S31 of FOI), and they have powers to enforce that decision. Aside from Elizabeth Denham’s obsession with data analytics in politics (especially when allegedly deployed by the Leave side), the ICO does not have a strong track record of taking on big organisations. Admittedly, the ICO recently took on the Metropolitan Police over their Gangs Matrix database, but the problem with that is the Gangs Matrix was a mess and the Met more or less acknowledged that.

The problem here is that if the Home Office maintain their position, the ICO would have to substitute their judgment for their’s. This wouldn’t be a mistake or a cock-up; if the Home Office use the DPA exemptions in the same way as they have the FOI ones, the only way that people can get better transparency is for the ICO to tell them that they’re wrong. This is often when Wilmslow bottles it. It’s straightforward to enforce on an organisation that has just lost thousands of people’s data (I’m sure it takes a lot of graft, but the decision to do it isn’t as hard). It’s much more difficult when the data controller hasn’t made a mistake, but is using the exemptions as described. Even if the ICO believes that the exemptions have been wrongly applied (and they might not), the Home Office is likely to ignore any recommendations and appeal any enforcement action.

The alternative is the courts, which is just as much of a roll of the dice as a complaint to the ICO, with the added complexity and cost of actually going to court. I have confidence that a court would test the Home Office’s arguments more robustly that the ICO would, but the Home Office wouldn’t be acting irrationally or unreasonably, and a judge might agree with them. These exemptions made it through Parliament and are on the statute book; the Home Office can plainly use them, and it’s not a breach of the GDPR unless the ICO or a court says that they have been applied unfairly.

Personally, I doubt that knowing who is receiving your data will undermine this process sufficiently justify the secrecy that the Home Office has already imposed using FOI, and which I expect they will use under DP, but it doesn’t matter what I think. This is where the hype around the GDPR runs into the brick wall of reality. The Home Office doesn’t need consent to gather, use and disclose personal data in this process, as long as it has another lawful basis to do so (legal obligation or official authority will certainly kick in here). The DPA gives them exemptions to keep the nature of that processing opaque, and if they choose to use them, challenging that decision is difficult and the outcome is uncertain. This leaves an odd situation but a lawful one – if they wish to live in a country they have already made their home, it seems that EU citizens have to submit to a closed, secretive process and they cannot find out what happens to their data during that process, who gets to see it, and for what purpose.

We’ve had years of headlines about Cambridge Analytica and Facebook which have captured the public’s imagination like never before, and generated huge publicity for the Information Commissioner’s Office and their army of blue-jacketed enforcers. Action, on the other hand, has been slightly less forthcoming. No action has been taken against Cambridge Analytica itself – there is the prosecution of SCL Elections over a subject access request made by an American (David Carroll), but if anyone can explain why prosecuting the now defunct company when the best outcome is a fine that will never be paid because it will be buried at the bottom of the pile of creditors, comment below. The ICO issued their first GDPR enforcement notice against AIQ, and it was so clumsy it had to be withdrawn and replaced (it’s astonishing that the ICO’s mishandling of this landmark action has gone virtually unnoticed). There is the famous Facebook fine of course, but that is already under appeal. Given that the Commissioner’s case changed radically from the Notice of Intent (published against all normal ICO practice) to final penalty, I don’t think that the ICO should count any chickens on the outcome.

The other issue haunting the case is a number of legal firms mounting ambitious compensation claims on behalf of those who believe themselves to be affected. Just as I am sceptical about the ICO’s track record, some odd assertions in a story in the Independent about David Carroll’s own attempt to sue Cambridge Analytica make me wonder whether the compensation road will be any less rocky. The claim is happening under the old Data Protection Act, and so Carroll and his solicitors will have to prove some kind of damage. Carroll’s solicitor Ravi Naik from ITN Solicitors is quoted as saying payouts could spiral to as much as £43 billion if only 10% of the possible affected pool of people claimed successfully.

“Even if one conservatively uses the lowest end of the range, both in number and value of each claim, and calculates on the basis of 10 per cent of the estimated 87 million affected Facebook users only, with claims of £5,000 each against Cambridge Analytica, that still implies a total potential claim value of £43.5bn”

I think his claims are optimistic at best, and at worst, comically exaggerated. Facebook did claim that up to 87 million people’s data may have been affected, but they’ve wavered since – to the extent that the ICO now admit that UK data wasn’t used by Cambridge Analytica in their final penalty on Facebook, despite building their NOI around that very claim. Carroll is claiming between £5000 and £20000, but he won’t get a penny unless he can show evidence of the breach in the first place, and then evidence of the damage. Claiming compensation for non-material damage is tricky. You can’t show something concrete like lost wages or business – the money won’t be awarded just because Carroll says he’s upset or annoyed, and the courts have shown scepticism in the past about claims of damage or distress (look at the Tetrus case that ICO lost on the issue of distress a few years back).

That 87 million number is a maximum, not a certainty, and the UK courts have shown themselves to be unmoved by generic class action claims of damage. Look at Richard Lloyd’s failed claim against Google, where the court said that different people will react to the use of their data in different ways. Perhaps Carroll has made a good case about the harm he says was done to him, but even if he has, that is not to say that all claimants are in the same position. If my data was abused by Facebook, my reaction would be numb resignation at worst. I can’t get outraged about Facebook abusing my data, any more than I can get upset by rain being wet. This is why I don’t use Facebook.

The consensus on LinkedIn seems to be that a possible breach is automatically accompanied by a ringing cash register – but that’s not a safe assumption, backed by any evidence. Lloyd lost his Google claim. Everyone who wrote excited Tweets and LinkedIn posts about the outcome of the recent Morrisons case – where the supermarket was found vicariously liable for a breach committed by an employee – ignored the fact that even if Morrisons lose their planned appeal to the Supreme Court, the issue of how much each claimant gets hasn’t been considered yet. Admittedly, Morrisons is a claim for misuse of private information and breach of confidence, but even so, we haven’t got to the bit about the money yet. The claimants may each get a big payout; they may get bus fare. There hasn’t been a case in the UK where multiple people received a big payout because their personal data was abused.

Naik’s extravagant claims and ambitious maths make for an impressive headline, but it’s speculation. I’m uncomfortable about the idea of tempting people into joining litigation (which is presumably the point of Naik’s claim) using hyped-up numbers in this way. The words sound sensible, and Naik effectively describes his estimate as conservative, but it’s a fantasy. Carroll will lose unless he can persuade the court that a breach occurred, that he experienced damage, and that there is a figure that will compensate him for that harm. We have had a few interesting and successful compensation claims in the past, but the idea that we’re looking at lottery jackpots for DP claimants is, so far, Fake News.