Category Archives: National Pupil Database

Data Protection law is being set up as a patsy.

After Dominic Cummings’ marathon session at the Select Committee, the Times published an article on,”The heroes and villains of the pandemic, according to Dominic Cummings”

One of Dom’s villains left out, was data protection law. He claimed, “if someone somewhere in the system didn’t say, ‘ignore GDPR’ thousands of people were going to die,” and that “no one even knew if that itself was legal—it almost definitely wasn’t.”

Thousands of people have died since that event he recalled from March 2020, but as a result of Ministers’ decisions, not data laws.

Data protection laws are *not* barriers, but permissive laws to *enable* use of personal data within a set of standards and safeguards designed to protect people. The opposite of what its detractors would have us believe.

The starting point is fundamental human rights. Common law confidentially. But the GDPR and its related parts on public health, are in fact specifically designed to enable data processing that overrules those principles for pandemic response purposes . In recognition of emergency needs for a limited time period, data protection laws permit interference with our fundamental rights and freedoms, including overriding privacy.

We need that protection of our privacy sometimes from government itself. And sometimes from those who see themselves as “the good guys” and above the law.

The Department of Health appears to have no plan to tell people about care.data 2,  the latest attempt at an NHS data grab, despite the fact that data protection laws require that they do. From September 1st (delayed to enable it to be done right, thanks to campaign efforts from medConfidential et supporters) all our GP medical records will be copied into a new national database for re-use, unless we actively opt out.

It’s groundhog day for the Department of Health. It is baffling why the government cannot understand or accept the need to do the right thing, and instead is repeating the same mistake of recent memory, all over again. Why the rush without due process and steamrollering any respect for the rule of law?

Were it not so serious, it might amuse me that some academic researchers appear to fail to acknowledge this matters, and they are getting irate on Twitter that *privacy* or ‘campaigners’ will prevent them getting hold of the data they appear to feel entitled to. Blame the people that designed a policy that will breach human rights and the law, not the people who want your rights upheld. And to blame the right itself is just, frankly, bizarre.

Such rants prompt me to recall the time when early on in my lay role on the Administrative Data Research Network approvals panel, a Director attending the meeting *as a guest* became so apoplectic with rage, that his face was nearly purple. He screamed, literally, at the panel of over ten well respected academics and experts in research and / or data because he believed the questions being asked over privacy and ethics principles in designing governance documents were unnecessary.

Or I might recall the request at my final meeting two years later in 2017 by another then Director, for access to highly sensitive and linked children’s health and education data to do (what I believed was valuable) public interest research involving the personal data of children with Down Syndrome. But the request came through the process with no ethical review. A necessary step before it should even have reached the panel for discussion.

I was left feeling from those two experiences, that both considered themselves and their work to be in effect “above the law” and expected special treatment, and a free pass without challenge. And that it had not improved over the two years.

If anyone in the research community cannot support due process, law, and human rights when it comes to admin data access, research using highly sensitive data about people’s lives with potential for significant community and personal impacts, then you are part of the problem.  There was extensive public outreach in 2012-13 across the UK about the use of personal if de-identified data in safe settings. And in 2014 the same concerns and red-lines were raised by hundreds of people in person, almost universally with the same reactions, at a range of care.data public engagement events. Feedback which institutions say matters, but continue to ignore.

It seems nothing has changed since I wrote,

“The commercial intermediaries still need to be told, don’t pee in the pool. It spoils it, for everyone else.”

We could also look back to when Michael Gove as Secretary of State for Education, changed the law in 2012 to permit pupil level, identifying and sensitive personal data to be given away to third parties. Press. Charities. Commercial companies even included an Internet tutoring business. (Cummings was coincidentally his SpAd at the Department for Education.)  As a direct result of that decision to give away pupils’ personal data in 2012, (in effect ‘re-engineering’ how the education sector was structured and the roles of the local authority and non-state providers)  an ICO audit of the DfE in February 2020 found unlawful practice and made 139 recommendations for change. We’re still waiting to see if and how it will be fixed.  At the moment it’s business as usual. Literally. The ICO don’t appear even to have stopped further data distribution until made lawful.

In April 2021, in answer to a written Parliamentary Question Nick Gibb, Schools Minister, made a commitment to “publish an update to the audit in June 2021 and further details regarding the release mechanism of the full audit report will be contained in this update.”  Will they promote openess, transparency, accountablity,or continue to skulk from publishing the whole truth?

Children have lost control of their digital footprint in state education by their fifth birthday.  The majority of parents polled in 2018 do not know the National Pupil Database even exists. 69% of over 1,004 parents asked, replied that they had not been informed that the Department for Education may give away children’s data to third-parties at all.

Thousands of companies continue to exploit children’s school records, without opt-in or opt-out, including special educational needs, ethnicity, and other sensitive data at pupil level.

Data protection law alone is in fact so enabling of data flow, that it is inadequate to protect children’s rights and freedoms across the state education sector in England; whether from public interest, charity or commercial research interventions without opt in or out, without parental knowledge. We shouldn’t need to understand our rights or to be proactive, in order to have them protected by default but data protection law and the ICO in particular have been captured by the siren call of data as a source of ‘innovation’ and economic growth.

Throughout 2018 and questions over Vote Leave data uses, Cummings claimed to know GDPR well. It was everyone else who didn’t. On his blog that July he suggested, “MPs haven’t even bothered to understand GDPR, which they mis-explain badly,” and in April he wrote,  The GDPR legislation is horrific. One of the many advantages of Brexit is we will soon be able to bin such idiotic laws.” He lambasted the Charter of Fundamental Rights the protections of which the government went on to take away from us under European Union Withdrawal Act.

But suddenly, come 2020/21 he is suggesting he didn’t know the law that well after all, “no one even knew if that itself was legal—it almost definitely wasn’t.”

Data Protection law is being set up as a patsy, while our confidentiality is commodified. The problem is not the law. The problem is those in power who fail to respect it, those who believe themselves to be above it, and who feel an entitlement to exploit that for their own aims.


Added 21/06/2021: Today I again came across a statement that I thought worth mentioning, from the Explanatory Notes for the Data Protection Bill from 2017:

Accordingly, Parliament passed the Data Protection Act 1984 and ratified the Convention in 1985, partly to ensure the free movement of data. The Data Protection Act 1984 contained principles which were taken almost directly from Convention 108 – including that personal data shall be obtained and processed fairly and lawfully and held only for specified purposes.”

The Data Protection Directive (95/46/EC) (“the 1995 Directive”) provides the current basis for the UK’s data protection regime. The 1995 Directive stemmed from the European Commission’s concern that a number of Member States had not introduced national law related to Convention 108 which led to concern that barriers may be erected to data flows. In addition, there was a considerable divergence in the data protection laws between Member States. The focus of the 1995 Directive was to protect the right to privacy with respect to the processing of personal data and to ensure the free flow of personal data between Member States. “

The National Data Strategy. Rewiring State power.

The National Data Strategy is not about “the data”.  People need to stop thinking of data only as abstract information, or even as personal data when it comes to national policy. Administrative data is the knowledge about the workings of the interactions between the public and the State and about us as people. It is the story of selected life events. To the State it is business intelligence. What resources are used, where, by whom and who costs The Treasury how much? How any government is permitted to govern that,  shapes our relationship with the State and the nature of the governance we get of people, of public services. How much power we cede to the State or retain at national, local, and individual levels over our communities and our lives matters.  Any change in National Data Strategy is about the rewiring of state power, and we need to understand its terms and conditions very, very carefully.


What government wants

“It’s not to say we don’t see privacy and data protection as not important,” said Phil Earl, Deputy Director at DCMS, in the panel discussion hosted by techUK as part of Birmingham Tech Week, exploring the UK Government’s recently released National Data Strategy.

I sighed so loudly I was glad to be on mute. The first of many big watch outs for the messaging around the National Data Strategy was already touted in the text, as “the high watermark of data use set during the pandemic.” In response to COVID “a few of the perceived barriers seem to have melted away,” said Earl, and saw this reduced state of data protections is desirable beyond the pandemic. “Can we maintain that level of collaboration and willingness to share data?” he asked.

Data protection laws are at their heart protections for people, not data, and if any government is seeking to reduce those protections for people we should pay attention to messaging very carefully.

This positioning fails to recognise that data protection law is more permissive in exceptional circumstances such as pandemics, with a recognition by default that the tests in law of necessity and proportionality are different from usual, and are time bound to the pandemic.

“What’s the equivalent? How do we empower people to feel that that greater good [felt in the pandemic] outweighs their legitimate concerns about data being shared,” he said.” The whole trust thing is something that must be constantly maintained,” but you may hear between the lines,  ‘preferably on our [government] terms.’

The idea that the public is ignorant about data, is often repeated and still wrong. The same old mantras resurfaced. If people can make more informed decisions, understand “the benefits”, then the government can influence their judgments, trusting us to “make the decisions that we want them to make [to agree to data re-use].”

If *this* is the government set course (again), then watch out.

What people want

In fact when asked, the majority of people both who are willing and less willing to have data about them reused, generally want the same things. Safeguards,  opt in to re use, restricted distribution, and protections for redress and against misuse strengthened in legislation.

Read Doteveryone’s public attitudes work. Or the Ipsos MORI polls or work by Wellcome. (see below). Or even the care.data summaries.

The red lines in the report from workshops carried out across different regions of the UK for the 2013 ADRN (about reuse of deidentified linked public admin datasets by qualified researchers in safe settings), remain valid today, in particular with relation to:

  • Creating large databases containing many variables/data from a large number of public sector sources

  • Allowing administrative data to be linked with business data

  • Linking of passively collected administrative data, in particular geo-location data

“All of the above were seen as having potential privacy implications or allowing the possibility of reidentification of individuals within datasets. The other ‘red-line’ for some participants was allowing researchers for private companies to access data, either to deliver a public service or in order to make profit. Trust in private companies’ motivations were low.”

The BT spokesperson on the panel, went on to say that their own survey showed 85% of people say their data is important to them, and 75% believe they have too little control.

Mr. Earl was absolutely correct in saying it puts the onus on government to be transparent and show how data will be used. But we hear *nothing* about concrete plans to deliver that. What does that look like? Saying it three times out loud, doesn’t make it real.

What government does

Government needs to embrace the fact it can only get data right, if it does the right thing. That includes upholding the law. This includes examining its own purposes and practice.

The Department for Education has been giving away 15 million people’s personal confidential data since 2012 and never told them. They knew this. They chose to ignore it. And on top of that, didn’t inform people who were in school since then, that Mr Gove changed the law. So now over 21 million people’s pupil records are being given away to comapnies and other third parties, for use in ways we do not expect, and is misused too. In 2015, more secret data sharing began, with the Home Office. And another pilot in 2018 with the DWP.  And in 2019, sharing with the police.

Promises on government data policy transparency right now are worth less than zero. What matters now is government actions. Trust will be based on what you do, not what you say. Is the government trustworthy?

After the summary findings published by the ICO of their compulsory audit of the Department for Education,  the question now is what will the Department and government do to address the 139 recommendations for improvement, with over 60% classified as urgent or high priority. Is the government intentional about change?

What will government do?

So I had a question for the panel: Is the government serious about its point in the strategy, 7.1.2 “Our data regime should empower individuals and groups to control and understand how their data is used.”

I didn’t get an answer.

I want to know if the government is prepared to build the necessary infrastructure to enable that understanding and control?

  • Enhance and build the physical infrastructure:
      • access to personal reports what data is held and how it is used.
      • management controls and communications over reuse [opt-in to meet the necessary permissions of legitimate interests or consent as lawful basis for further data processing, conditions for sensitive data processing, or at very least opt-out to respect objections].
      • secure systems (not just excel, and WannaCry resistant)
  • Enable the necessary transparency tools and create demonstrable accountability through registers of algorithms and data sharing with oversight functions and routes for redress.
  • Empower staff with the necessary human skills at all levels in the public sector on the laws around data that do not just consist of a sharepoint on GDPR — what about privacy law, communications law, equality and discrimination laws among others?
  • Empower the public with the controls they want to have their rights respected.
  • Examine toxic policy that drives bad data collection and re-use.

Public admin data is all about people

Coming soon is the publication of an Integrated Review, we were told, how ‘data and security’ and other joined up issues will feature.

A risk of this conflation is seeing the national data strategy as another dry review about data as ‘a thing’, or its management.

It should be about people. The people who our public admin data are about. The people that want access to it. The people making the policy decisions. And its human infrastructure. The controls on power about purposes of the data reuse, data governance is all about the roles and responsibilties of people and the laws that oversee them and require human accountability.

These NDS strategy missions, and pillars and aims are all about “the data”.

For a national data strategy to be realised and to thrive in all of our wide ranging societal ‘data’ interests,  it cannot be all about data as a commodity.  Or all about government wants. Or seen through the lens only of research. Allow that, and they divide and conquer. It must be founded on  building a social contract between government and the public in a digital environment, and setting the expectations of these multi-purpose relationships, at national, and local levels.

For a forward thinking data strategy truly building something in the long term public interest, it is  about understanding ‘wider public need‘. The purpose of ‘the data’ and its strategy, is as much about the purpose of government behind it. Data is personal. But when used to make decisions it also business intelligence. How does the government make the business of governing work, through data?

Any national data strategy does not sit in a vacuum of other policy and public experience of government either.  If Manchester‘s second lockdown funding treatment is seen as the expectations of how local needs get trumped by national wants, and people’s wishes will be steam rollered, then a national approach will not win support. A long list of bad government engagement over recent months, is a poor foundation and you don’t fix that by talking about “the benefits”.

Will government listen?

Edgenuity, the U.S. based school assessment system using AI for marking, made the news this summer, when parents found it could be gamed by simply packing essays with all the right keywords, but respondents didn’t need to make sense or give accurate answers. To be received well and get a good grade, they were expected simply to tell the system the words ‘it wanted to hear’.

If the teachers were looking at the responses, they didn’t care,” one student said.

Will the government actually look at responses to the National Data Strategy and care about getting it right? Not just care about getting what they want? Or about what commercial actors will do with it?

Government wanted to and changed the law on education admin data in 2012 and got it wrong. Education data alone is a sin bin of bad habits and complete lack of public and professional engagement, before even starting to address data quality and accuracy and backwards looking policy built on bad historic data.

The Commercial department do not have appropriate controls in place to protect personal data being processed on behalf of the DfE by data processors.” (ICO audit of the DfE , 2020)

Gambling companies ended up misusing learner records.

Government wanted data from one Department to be collected for the purposes of another and got it wrong. People boycotted the collection until it was killed off.

Government changed the law on Higher Education in 2017 and got it wrong.  Now  third parties pass around named equality monitoring records like religion, sexual orientation, and disability and it is stored forever on named national pupil records. The Department for Education (DfE) now holds sexual orientation data on almost 3.2 million, and religious belief data on 3.7 million people.

What could possibly go wrong?

If the current path is any indicator, this government is little interested in local power, or people, and certainly not in our human rights. They are interested in centralised power. We should be very cautious about giving that all away to the State on its own terms.


 

The national data strategy consultation is open for submissions until

Samples of public engagement on data reuse

 

 

What happens when a Regulator doesn’t regulate

The news is full of the exam Regulator Ofqual right now, since yesterday’s A-Level results came out. In the outcry over the clear algorithmic injustice and inexplicable data-driven results, the data regulator, the Information Commissioner (ICO) remains silent.**

I have been told the Regulators worked together from early on in the process. So did this collaboration help or hinder the thousands of students and children whose rights the Regulators are supposed to work to protect?

I have my doubts, and here is why.

My child’s named national school records

On April 29, 2015 I wrote to the Department for Education (DfE) to ask for a copy of the data that they held about my eldest child in the National Pupil Database (NPD). A so-called Subject Access Request. The DfE responded on 12 May 2015 and refused, claiming an exemption, section 33(4) of the Data Protection Act 1998. In effect saying it was a research-only, not operational database.

Despite being a parent of three children in state education in England, there was no clear information available to me what the government held in this database about my children. Building on what others in civil society had done before, I began research into what data was held. From where it was sourced and how often it was collected. Who the DfE gave the data to. For what purposes. How long it was kept. And I discovered a growing database of over 20 million individuals, of identifying and sensitive personal data, that is given away to commercial companies, charities, think tanks and press without suppression of small numbers and is never destroyed.

My children’s confidential records that I entrusted to school, and much more information they create that I never see, is given away for commercial purposes and I don’t get told which companies have it, why, or have any control over it? What about my Right to Object? I had imagined a school would only share statistics with third parties without parents’ knowledge or being asked. Well that’s nothing compared with what the Department does with it all next.

My 2015 complaint to the ICO

On October 6, 2015 I made a complaint to the Information Commissioner’s Office (the ICO). Admittedly, I was more naïve and less well informed than I am today, but the facts were clear.

Their response in April 2016, was to accept the DfE position, “at the stage at which this data forms part of its evidence base for certain purposes, it has been anonymised, aggregated and is statistical in nature.  Therefore, for the purposes of the DPA, at the stage at which the DfE use NPD data for such purposes, it no longer constitutes personal data in any event.”

The ICO was “satisfied that the DfE met the criteria needed to rely on the exemption contained at section 33(4) of the DPA” and was justified in not fulfilling my request.

And “in relation to your concerns about the NPD and the adequacy of the privacy notice provided by the DfE, in broad terms, we consider it likely that this complies with the relevant data protection principles of the DPA.”

The ICO claimed “the processing does not cause any substantial damage or distress to individuals and that any results of the research/statistics are not made available in a form which identifies data subjects.”

The ICO kept its eyes wide shut

In secret in July 2015, the DfE had started to supply the Home Office with the matched personal details of children from the NPD, including home address. The Home Office requested this for purposes including to further the Hostile Environment. (15.1.2) which I only discovered in detail one year to the day after my ICO complaint, on October 6, 2016. The rest is public.

Had the ICO investigated the uses of national pupil data a year earlier in 2015-16, might it have prevented this ongoing gross misuse of children’s personal data and public and professional trust?

The ICO made no public statement despite widespread media coverage throughout 2016 and legal action on the expansion of the data, and intended use of children’s  nationality and country-of-birth.

Identifying and sensitive not aggregated and  statistical

Since 2012 the DfE has given away the sensitive and identifying personal confidential data of over 23 million people without their knowledge, in over 1600 unique requests, that are not anonymous.

In 2015 there was no Data Protection Impact Assessment. The Department had done zero audits of data users after sending them identifying pupil data. There was no ethics process or paperwork.

Today in England pupil data are less protected than across the rest of the UK. The NPD is being used as a source for creating a parent polling panel. Onward data sharing is opaque but some companies continue to receive named data and have done so for many years. It is a linked dataset with over 25 different collections, and includes highly sensitive children’s social care data, is frequently expanded, its content scope grows increasiningly sensitive, facilitates linkage to external datasets including children at risk and for policing,  and has been used in criminology interventions which did harm and told children they were involved because they were “the worst kids.” Data has been given to journalists and a tutoring company. It has been sought after by Americans even if not (yet?) given to them.

Is the ICO a help or hindrance to protect children and young people’s data rights?

Five years ago the ICO told me the named records in the national pupil database was not personal data. Five years on, my legal team and I await a final regulatory response from the ICO that I hope will protect the human rights of my children, the millions currently in education in England whose data are actively collected, the millions aged 18-37 affected whose data were collected 1996-2012 but who don’t know, and those to come.

It has come at significant personal and legal costs and I welcome any support. It must be fixed. The question is whether the information rights Regulator is a help or hindrance?

If the ICO is working with organisations that have broken the law, or that plan dubious or unethical data processing, why is the Regulator collaborating to enable processing and showing them how to smooth off the edges rather than preventing harm and protecting rights? Can the ICO be both a friend and independent enforcer?

Why does it decline to take up complaints on behalf of parents that similarly affect millions of children in the UK and worldwide about companies that claim to use AI on their website but tell the ICO it’s just computing really. Or why has it given a green light on the reuse of religion and ethnicity from schools without consent, and tells the organisation they can later process it to make it anonymous, and keep all the personal data indefinitely?

I am angry at their inaction, but not as angry as thousands of children and their parents who know they have been let down by  data-led decisions this month, that to them are inexplicable.

Thousands of children who are caught up in the algorithmic A-Level debacle and will be in next week’s GCSE processes believe they have been unfairly treated through the use of their personal data and have no clear route of redress. Where is the voice of the Regulator? What harm should they have prevented but didn’t through inaction?

What is the point of all the press and posturing on an Age Appropriate Code of Practice which goes beyond the scope of data protection, if the ICO cannot or will not enforce on its core remit or support the public it is supposed to serve?


Update: This post was published at midday on Friday Aust 14. In the late afternoon the ICO did post a short statement on the A-levels crisis, and also wrote to me regarding one of these cases via email.

Can Data Trusts be trustworthy?

The Lords Select Committee report on AI in the UK in March 2018, suggested that,“the Government plans to adopt the Hall-Pesenti Review recommendation that ‘data trusts’ be established to facilitate the ethical sharing of data between organisations.”

Since data distribution already happens, what difference would a Data Trust model make to ‘ethical sharing‘?

A ‘set of relationships underpinned by a repeatable framework, compliant with parties’ obligations’ seems little better than what we have today, with all its problems including deeply unethical policy and practice.

The ODI set out some of the characteristics Data Trusts might have or share. As importantly, we should define what Data Trusts are not. They should not simply be a new name for pooling content and a new single distribution point. Click and collect.

But is a Data Trust little more than a new description for what goes on already? Either a physical space or legal agreements for data users to pass around the personal data from the unsuspecting, and sometimes unwilling, public. Friends-with-benefits who each bring something to the party to share with the others?

As with any communal risk, it is the standards of the weakest link, the least ethical, the one that pees in the pool, that will increase reputational risk for all who take part, and spoil it for everyone.

Importantly, the Lords AI Committee report recognised that there is an inherent risk how the public would react to Data Trusts, because there is no social license for this new data sharing.

“Under the current proposals, individuals who have their personal data contained within these trusts would have no means by which they could make their views heard, or shape the decisions of these trusts.

Views those keen on Data Trusts seem keen to ignore.

When the Administrative Data Research Network was set up in 2013, a new infrastructure for “deidentified” data linkage, extensive public dialogue was carried across across the UK. It concluded in very similar findings as was apparent at dozens of care.data engagement events in 2014-15;

There is not public support for

  • “Creating large databases containing many variables/data from a large number of public sector sources,
  • Establishing greater permanency of datasets,
  • Allowing administrative data to be linked with business data, or
  • Linking of passively collected administrative data, in particular geo-location data”

The other ‘red-line’ for some participants was allowing “researchers for private companies to access data, either to deliver a public service or in order to make profit. Trust in private companies’ motivations were low.”

All of the above could be central to Data Trusts. All of the above highlight that in any new push to exploit personal data, the public must not be the last to know. And until all of the above are resolved, that social-license underpinning the work will always be missing.

Take the National Pupil Database (NPD) as a case study in a Data Trust done wrong.

It is a mega-database of over 20 other datasets. Raw data has been farmed out for years under terms and conditions to third parties, including users who hold an entire copy of the database, such as the somewhat secretive and unaccountable Fischer Family Trust, and others, who don’t answer to Freedom-of-Information, and whose terms are hidden under commercial confidentilaity. Buying and benchmarking data from schools and selling it back to some, profiling is hidden from parents and pupils, yet the FFT predictive risk scoring can shape a child’s school experience from age 2. They don’t really want to answer how staff tell if a child’s FFT profile and risk score predictions are accurate, or of they can spot errors or a wrong data input somewhere.

Even as the NPD moves towards risk reduction, its issues remain. When will children be told how data about them are used?

Is it any wonder that many people in the UK feel a resentment of institutions and orgs who feel entitled to exploit them, or nudge their behaviour, and a need to ‘take back control’?

It is naïve for those working in data policy and research to think that it does not apply to them.

We already have safe infrastructures in the UK for excellent data access. What users are missing, is the social license to do so.

Some of today’s data uses are ethically problematic.

No one should be talking about increasing access to public data, before delivering increased public understanding. Data users must get over their fear of what if the public found out.

If your data use being on the front pages would make you nervous, maybe it’s a clue you should be doing something differently. If you don’t trust the public would support it, then perhaps it doesn’t deserve to be trusted. Respect individuals’ dignity and human rights. Stop doing stupid things that undermine everything.

Build the social license that care.data was missing. Be honest. Respect our right to know, and right to object. Build them into a public UK data strategy to be understood and be proud of.


Part 1. Ethically problematic
Ethics is dissolving into little more than a buzzword. Can we find solutions underpinned by law, and ethics, and put the person first?

Part 2. Can Data Trusts be trustworthy?
As long as data users ignore data subjects rights, Data Trusts have no social license.



Information society services: Children in the GDPR, Digital Economy Bill & Digital Strategy

In preparation for The General Data Protection Regulation (GDPR) there  must be an active UK decision about policy in the coming months for children and the Internet – provision of ‘Information Society Services’. The age of consent for online content aimed at children from May 25, 2018 will be 16 by default unless UK law is made to lower it.

Age verification for online information services in the GDPR, will mean capturing parent-child relationships. This could mean a parent’s email or credit card unless there are other choices made. What will that mean for access to services for children and to privacy? It is likely to offer companies an opportunity for a data grab, and mean privacy loss for the public, as more data about family relationships will be created and collected than the content provider would get otherwise.

Our interactions create a blended identity of online and offline attributes which I suggested in a previous post, create synthesised versions of our selves raises questions on data privacy and security.

The goal may be to protect the physical child. The outcome will mean it simultaneously expose children and parents to risks that we would not otherwise be put through increased personal data collection. By increasing the data collected, it increases the associated risks of loss, theft, and harm to identity integrity. How will legislation balance these risks and rights to participation?

The UK government has various work in progress before then, that could address these questions:

But will they?

As Sonia Livingstone wrote in the post on the LSE media blog about what to expect from the GDPR and its online challenges for children:

“Now the UK, along with other Member States, has until May 2018 to get its house in order”.

What will that order look like?

The Digital Strategy and Ed Tech

The Digital Strategy commits to changes in National Pupil Data  management. That is, changes in the handling and secondary uses of data collected from pupils in the school census, like using it for national research and planning.

It also means giving data to commercial companies and the press. Companies such as private tutor pupil matching services, and data intermediaries. Journalists at the Times and the Telegraph.

Access to NPD via the ONS VML would mean safe data use, in safe settings, by safe (trained and accredited) users.

Sensitive data — it remains to be seen how DfE intends to interpret ‘sensitive’ and whether that is the DPA1998 term or lay term meaning ‘identifying’ as it should — will no longer be seen by users for secondary uses outside safe settings.

However, a grey area on privacy and security remains in the “Data Exchange” which will enable EdTech products to “talk to each other”.

The aim of changes in data access is to ensure that children’s data integrity and identity are secure.  Let’s hope the intention that “at all times, the need to preserve appropriate privacy and security will remain paramount and will be non-negotiable” applies across all closed pupil data, and not only to that which may be made available via the VML.

This strategy is still far from clear or set in place.

The Digital Strategy and consumer data rights

The Digital Strategy commits under the heading of “Unlocking the power of data in the UK economy and improving public confidence in its use” to the implementation of the General Data Protection Regulation by May 2018. The Strategy frames this as a business issue, labelling data as “a global commodity” and as such, its handling is framed solely as a requirements needed to ensure “that our businesses can continue to compete and communicate effectively around the world” and that adoption “will ensure a shared and higher standard of protection for consumers and their data.”

The GDPR as far as children goes, is far more about protection of children as people. It focuses on returning control over children’s own identity and being able to revoke control by others, rather than consumer rights.

That said, there are data rights issues which are also consumer issues and  product safety failures posing real risk of harm.

Neither The Digital Economy Bill nor the Digital Strategy address these rights and security issues, particularly when posed by the Internet of Things with any meaningful effect.

In fact, the chapter Internet of Things and Smart Infrastructure [ 9/19]  singularly miss out anything on security and safety:

“We want the UK to remain an international leader in R&D and adoption of IoT. We are funding research and innovation through the three year, £30 million IoT UK Programme.”

There was much more thoughtful detail in the 2014 Blackett Review on the IoT to which I was signposted today after yesterday’s post.

If it’s not scary enough for the public to think that their sex secrets and devices are hackable, perhaps it will kill public trust in connected devices more when they find strangers talking to their children through a baby monitor or toy. [BEUC campaign report on #Toyfail]

“The internet-connected toys ‘My Friend Cayla’ and ‘i-Que’ fail miserably when it comes to safeguarding basic consumer rights, security, and privacy. Both toys are sold widely in the EU.”

Digital skills and training in the strategy doesn’t touch on any form of change management plans for existing working sectors in which we expect to see machine learning and AI change the job market. This is something the digital and industrial strategy must be addressing hand in glove.

The tactics and training providers listed sound super, but there does not appear to be an aspirational strategy hidden between the lines.

The Digital Economy Bill and citizens’ data rights

While the rest of Europe in this legislation has recognised that a future thinking digital world without boundaries, needs future thinking on data protection and empowered citizens with better control of identity, the UK government appears intent on taking ours away.

To take only one example for children, the Digital Economy Bill in Cabinet Office led meetings was explicit about use for identifying and tracking individuals labelled under “Troubled Families” and interventions with them. Why, when consent is required to work directly with people, that consent is being ignored to access their information is baffling and in conflict with both the spirit and letter of GDPR. Students and Applicants will see their personal data sent to the Student Loans Company without their consent or knowledge. This overrides the current consent model in place at UCAS.

It is baffling that the government is pursuing the Digital Economy Bill data copying clauses relentlessly, that remove confidentiality by default, and will release our identities in birth, marriage and death data for third party use without consent through Chapter 2, the opening of the Civil Registry, without any safeguards in the bill.

Government has not only excluded important aspects of Parliamentary scrutiny in the bill, it is trying to introduce “almost untrammeled powers” (paragraph 21), that will “very significantly broaden the scope for the sharing of information” and “specified persons”  which applies “whether the service provider concerned is in the public sector or is a charity or a commercial organisation” and non-specific purposes for which the information may be disclosed or used. [Reference: Scrutiny committee comments]

Future changes need future joined up thinking

While it is important to learn from the past, I worry that the effort some social scientists put into looking backwards,  is not matched by enthusiasm to look ahead and making active recommendations for a better future.

Society appears to have its eyes wide shut to the risks of coercive control and nudge as research among academics and government departments moves in the direction of predictive data analysis.

Uses of administrative big data and publicly available social media data for example, in research and statistics, needs further new regulation in practice and policy but instead the Digital Economy Bill looks only at how more data can be got out of Department silos.

A certain intransigence about data sharing with researchers from government departments is understandable. What’s the incentive for DWP to release data showing its policy may kill people?

Westminster may fear it has more to lose from data releases and don’t seek out the political capital to be had from good news.

The ethics of data science are applied patchily at best in government, and inconsistently in academic expectations.

Some researchers have identified this but there seems little will to action:

 “It will no longer be possible to assume that secondary data use is ethically unproblematic.”

[Data Horizons: New forms of Data for Social Research, Elliot, M., Purdam, K., Mackey, E., School of Social Sciences, The University Of Manchester, 2013.]

Research and legislation alike seem hell bent on the low hanging fruit but miss out the really hard things. What meaningful benefit will it bring by spending millions of pounds on exploiting these personal data and opening our identities to risk just to find out whether X course means people are employed in Y tax bracket 5 years later, versus course Z where everyone ends up self employed artists? What ethics will be applied to the outcomes of those questions asked and why?

And while government is busy joining up children’s education data throughout their lifetimes from age 2 across school, FE, HE, into their HMRC and DWP interactions, there is no public plan in the Digital Strategy for the coming 10 to 20 years employment market, when many believe, as do these authors in American Scientific, “around half of today’s jobs will be threatened by algorithms. 40% of today’s top 500 companies will have vanished in a decade.”

What benefit will it have to know what was, or for the plans around workforce and digital skills list ad hoc tactics, but no strategy?

We must safeguard jobs and societal needs, but just teaching people to code is not a solution to a fundamental gap in what our purpose will be, and the place of people as a world-leading tech nation after Brexit. We are going to have fewer talented people from across the world staying on after completing academic studies, because they’re not coming at all.

There may be investment in A.I. but where is the investment in good data practices around automation and machine learning in the Digital Economy Bill?

To do this Digital Strategy well, we need joined up thinking.

Improving online safety for children in The Green Paper on Children’s Internet Safety should mean one thing:

Children should be able to use online services without being used and abused by them.

This article arrived on my Twitter timeline via a number of people. Doteveryone CEO Rachel Coldicutt summed up various strands of thought I started to hear hints of last month at #CPDP2017 in Brussels:

“As designers and engineers, we’ve contributed to a post-thought world. In 2017, it’s time to start making people think again.

“We need to find new ways of putting friction and thoughtfulness back into the products we make.” [Glanceable truthiness, 30.1.2017]

Let’s keep the human in discussions about technology, and people first in our products

All too often in technology and even privacy discussions, people have become ‘consumers’ and ‘customers’ instead of people.

The Digital Strategy may seek to unlock “the power of data in the UK economy” but policy and legislation must put equal if not more emphasis on “improving public confidence in its use” if that long term opportunity is to be achieved.

And in technology discussions about AI and algorithms we hear very little about people at all.  Discussions I hear seem siloed instead into three camps: the academics, the designers and developers,  the politicians and policy makers.  And then comes the lowest circle, ‘the public’ and ‘society’.

It is therefore unsurprising that human rights have fallen down the ranking of importance in some areas of technology development.

It’s time to get this house in order.

Datasharing, lawmaking and ethics: power, practice and public policy

“Lawmaking is the Wire, not Schoolhouse Rock. It’s about blood and war and power, not evidence and argument and policy.”

"We can't trust the regulators," they say. "We need to be able to investigate the data for ourselves." Technology seems to provide the perfect solution. Just put it all online - people can go through the data while trusting no one.  There's just one problem. If you can't trust the regulators, what makes you think you can trust the data?" 

Extracts from The Boy Who Could Change the World: The Writings of Aaron Swartz. Chapter: ‘When is Technology Useful? ‘ June 2009.

The question keeps getting asked, is the concept of ethics obsolete in Big Data?

I’ve come to some conclusions why ‘Big Data’ use keeps pushing the boundaries of what many people find acceptable, and yet the people doing the research, the regulators and lawmakers often express surprise at negative reactions. Some even express disdain for public opinion, dismissing it as ignorant, not ‘understanding the benefits’, yet to be convinced. I’ve decided why I think what is considered ‘ethical’ in data science does not meet public expectation.

It’s not about people.

Researchers using large datasets, often have a foundation in data science, applied computing, maths, and don’t see data as people. It’s only data. Creating patterns, correlations, and analysis of individual level data are not seen as research involving human subjects.

This is embodied in the nth number of research ethics reviews I have read in the last year in which the question is asked, does the research involve people? The answer given is invariably ‘no’.

And these data analysts using, let’s say health data, are not working in a subject that is founded on any ethical principle, contrasting with the medical world the data come from.

The public feels differently about the information that is about them, and may be known, only to them or select professionals. The values that we as the public attach to our data  and expectations of its handling may reflect the expectation we have of handling of us as people who are connected to it. We see our data as all about us.

The values that are therefore put on data, and on how it can and should be used, can be at odds with one another, the public perception is not reciprocated by the researchers. This may be especially true if researchers are using data which has been de-identified, although it may not be anonymous.

New legislation on the horizon, the Better Use of Data in Government,  intends to fill the [loop]hole between what was legal to share in the past and what some want to exploit today, and emphasises a gap in the uses of data by public interest, academic researchers, and uses by government actors. The first incorporate by-and-large privacy and anonymisation techniques by design, versus the second designed for applied use of identifiable data.

Government departments and public bodies want to identify and track people who are somehow misaligned with the values of the system; either through fraud, debt, Troubled Families, or owing Student Loans. All highly sensitive subjects. But their ethical data science framework will not treat them as individuals, but only as data subjects. Or as groups who share certain characteristics.

The system again intrinsically fails to see these uses of data as being about individuals, but sees them as categories of people – “fraud” “debt” “Troubled families.” It is designed to profile people.

Services that weren’t built for people, but for government processes, result in datasets used in research, that aren’t well designed for research. So we now see attempts to shoehorn historical practices into data use  by modern data science practitioners, with policy that is shortsighted.

We can’t afford for these things to be so off axis, if civil service thinking is exploring “potential game-changers such as virtual reality for citizens in the autism spectrum, biometrics to reduce fraud, and data science and machine-learning to automate decisions.”

In an organisation such as DWP this must be really well designed since “the scale at which we operate is unprecedented: with 800 locations and 85,000  colleagues, we’re larger than most retail operations.”

The power to affect individual lives through poor technology is vast and some impacts seem to be being badly ignored. The ‘‘real time earnings’ database improved accuracy of benefit payments was widely agreed to have been harmful to some individuals through the Universal Credit scheme, with delayed payments meaning families at foodbanks, and contributing to worse.

“We believe execution is the major job of every business leader,” perhaps not the best wording in on DWP data uses.

What accountability will be built-by design?

I’ve been thinking recently about drawing a social ecological model of personal data empowerment or control. Thinking about visualisation of wants, gaps and consent models, to show rather than tell policy makers where these gaps exist in public perception and expectations, policy and practice. If anyone knows of one on data, please shout. I think it might be helpful.

But the data *is* all about people

Regardless whether they are in front of you or numbers on a screen, big or small datasets using data about real lives are data about people. And that triggers a need to treat the data with an ethical approach as you would people involved face-to-face.

Researchers need to stop treating data about people as meaningless data because that’s not how people think about their own data being used. Not only that, but if the whole point of your big data research is to have impact, your data outcomes, will change lives.

Tosh, I know some say. But, I have argued, the reason being is that the applications of the data science/ research/ policy findings / impact of immigration in education review / [insert purposes of the data user’s choosing] are designed to have impact on people. Often the people about whom the research is done without their knowledge or consent. And while most people say that is OK, where it’s public interest research, the possibilities are outstripping what the public has expressed as acceptable, and few seem to care.

Evidence from public engagement and ethics all say, hidden pigeon-holing, profiling, is unacceptable. Data Protection law has special requirements for it, on autonomous decisions. ‘Profiling’ is now clearly defined under article 4 of the GDPR as ” any form of automated processing of personal data consisting of using those data to evaluate certain personal aspects relating to a natural person, in particular to analyse or predict aspects concerning that natural person’s performance at work, economic situation, health, personal preferences, interests, reliability, behaviour, location or movements.”

Using big datasets for research that ‘isn’t interested in individuals’ may still intend to create results profiling groups for applied policing, or discriminate, to make knowledge available by location. The data may have been deidentified, but in application becomes no longer anonymous.

Big Data research that results in profiling groups with the intent for applied health policy impacts for good, may by the very point of research, with the intent of improving a particular ethnic minority access to services, for example.

Then look at the voting process changes in North Carolina and see how that same data, the same research knowledge might be applied to exclude, to restrict rights, and to disempower.

Is it possible to have ethical oversight that can protect good data use and protect people’s rights if they conflict with the policy purposes?

The “clear legal basis”is not enough for public trust

Data use can be legal and can still be unethical, harmful and shortsighted in many ways, for both the impacts on research – in terms of withholding data and falsifying data and avoiding the system to avoid giving in data – and the lives it will touch.

What education has to learn from health is whether it will permit the uses by ‘others’ outside education to jeopardise the collection of school data intended in the best interests of children, not the system. In England it must start to analyse what is needed vs wanted. What is necessary and proportionate and justifies maintaining named data indefinitely, exposed to changing scope.

In health, the most recent Caldicott review suggests scope change by design – that is a red line for many: “For that reason the Review recommends that, in due course, the opt-out should not apply to all flows of information into the HSCIC. This requires careful consideration with the primary care community.”

The community spoke out already, and strongly in Spring and Summer 2014 that there must be an absolute right to confidentiality to protect patients’ trust in the system. Scope that ‘sounds’ like it might sneakily change in future, will be a death knell to public interest research, because repeated trust erosion will be fatal.

Laws change to allow scope change without informing people whose data are being used for different purposes

Regulators must be seen to be trusted, if the data they regulate is to be trustworthy. Laws and regulators that plan scope for the future watering down of public protection, water down public trust from today. Unethical policy and practice, will not be saved by pseudo-data-science ethics.

Will those decisions in private political rooms be worth the public cost to research, to policy, and to the lives it will ultimately affect?

What happens when the ethical black holes in policy, lawmaking and practice collide?

At the last UK HealthCamp towards the end of the day, when we discussed the hard things, the topic inevitably moved swiftly to consent, to building big databases, public perception, and why anyone would think there is potential for abuse, when clearly the intended use is good.

The answer came back from one of the participants, “OK now it’s the time to say. Because, Nazis.” Meaning, let’s learn from history.

Given the state of UK politics, Go Home van policies, restaurant raids, the possibility of Trump getting access to UK sensitive data of all sorts from across the Atlantic, given recent policy effects on the rights of the disabled and others, I wonder if we would hear the gentle laughter in the room in answer to the same question today.

With what is reported as Whitehall’s digital leadership sharp change today, the future of digital in government services and policy and lawmaking does indeed seem to be more “about blood and war and power,” than “evidence and argument and policy“.

The concept of ethics in datasharing using public data in the UK is far from becoming obsolete. It has yet to begin.

We have ethical black holes in big data research, in big data policy, and big data practices in England. The conflicts between public interest research and government uses of population wide datasets, how the public perceive the use of our data and how they are used, gaps and tensions in policy and practice are there.

We are simply waiting for the Big Bang. Whether it will be creative, or destructive we are yet to feel.

*****

image credit: LIGO – graphical visualisation of black holes on the discovery of gravitational waves

References:

Report: Caldicott review – National Data Guardian for Health and Care Review of Data Security, Consent and Opt-Outs 2016

Report: The OneWay Mirror: Public attitudes to commercial access to health data

Royal Statistical Society Survey carried out by Ipsos MORI: The Data Trust Deficit

Can new datasharing laws win social legitimacy, public trust and support without public engagement?

I’ve been struck by stories I’ve heard on the datasharing consultation, on data science, and on data infrastructures as part of ‘government as a platform’ (#GaaPFuture) in recent weeks. The audio recorded by the Royal Statistical Society on March 17th is excellent, and there were some good questions asked.

There were even questions from insurance backed panels to open up more data for commercial users, and calls for journalists to be seen as accredited researchers, as well as to include health data sharing. Three things that some stakeholders, all users of data, feel are  missing from consultation, and possibly some of those with the most widespread public concern and lowest levels of public trust. [1]

What I feel is missing in consultation discussions are:

  1. a representative range of independent public voice
  2. a compelling story of needs – why tailored public services benefits citizens from whom data is taken, not only benefits data users
  3. the impacts we expect to see in local government
  4. any cost/risk/benefit assessment of those impacts, or for citizens
  5. how the changes will be independently evaluated – as some are to be reviewed

The Royal Statistical Society and ODI have good summaries here of their thoughts, more geared towards the statistical and research aspects of data,  infrastructure and the consultation.

I focus on the other strands that use identifiable data for targeted interventions. Tailored public services, Debt, Fraud, Energy Companies’ use. I think we talk too little of people, and real needs.

Why the State wants more datasharing is not yet a compelling story and public need and benefit seem weak.

So far the creation of new data intermediaries, giving copies of our personal data to other public bodies  – and let’s be clear that this often means through commercial representatives like G4S, Atos, Management consultancies and more –  is yet to convince me of true public needs for the people, versus wants from parts of the State.

What the consultation hopes to achieve, is new powers of law, to give increased data sharing increased legal authority. However this alone will not bring about the social legitimacy of datasharing that the consultation appears to seek through ‘open policy making’.

Legitimacy is badly needed if there is to be public and professional support for change and increased use of our personal data as held by the State, which is missing today,  as care.data starkly exposed. [2]

The gap between Social Legitimacy and the Law

Almost 8 months ago now, before I knew about the datasharing consultation work-in-progress, I suggested to BIS that there was an opportunity for the UK to drive excellence in public involvement in the use of public data by getting real engagement, through pro-active consent.

The carrot for this, is achieving the goal that government wants – greater legal clarity, the use of a significant number of consented people’s personal data for complex range of secondary uses as a secondary benefit.

It was ignored.

If some feel entitled to the right to infringe on citizens’ privacy through a new legal gateway because they believe the public benefit outweighs private rights, then they must also take on the increased balance of risk of doing so, and a responsibility to  do so safely. It is in principle a slippery slope. Any new safeguards and ethics for how this will be done are however unclear in those data strands which are for targeted individual interventions. Especially if predictive.

Upcoming discussions on codes of practice [which have still to be shared] should demonstrate how this is to happen in practice, but codes are not sufficient. Laws which enable will be pushed to their borderline of legal and beyond that of ethical.

In England who would have thought that the 2013 changes that permitted individual children’s data to be given to third parties [3] for educational purposes, would mean giving highly sensitive, identifiable data to journalists without pupils or parental consent? The wording allows it. It is legal. However it fails the DPA Act legal requirement of fair processing.  Above all, it lacks social legitimacy and common sense.

In Scotland, there is current anger over the intrusive ‘named person’ laws which lack both professional and public support and intrude on privacy. Concerns raised should be lessons to learn from in England.

Common sense says laws must take into account social legitimacy.

We have been told at the open policy meetings that this change will not remove the need for informed consent. To be informed, means creating the opportunity for proper communications, and also knowing how you can use the service without coercion, i.e. not having to consent to secondary data uses in order to get the service, and knowing to withdraw consent at any later date. How will that be offered with ways of achieving the removal of data after sharing?

The stick for change, is the legal duty that the recent 2015 CJEU ruling reiterating the legal duty to fair processing [4] waved about. Not just a nice to have, but State bodies’ responsibility to inform citizens when their personal data are used for purposes other than those for which those data had initially been consented and given. New legislation will not  remove this legal duty.

How will it be achieved without public engagement?

Engagement is not PR

Failure to act on what you hear from listening to the public is costly.

Engagement is not done *to* people, don’t think explain why we need the data and its public benefit’ will work. Policy makers must engage with fears and not seek to dismiss or diminish them, but acknowledge and mitigate them by designing technically acceptable solutions. Solutions that enable data sharing in a strong framework of privacy and ethics, not that sees these concepts as barriers. Solutions that have social legitimacy because people support them.

Mr Hunt’s promised February 2014 opt out of anonymised data being used in health research, has yet to be put in place and has had immeasurable costs for delayed public research, and public trust.

How long before people consider suing the DH as data controller for misuse? From where does the arrogance stem that decides to ignore legal rights, moral rights and public opinion of more people than those who voted for the Minister responsible for its delay?

 

This attitude is what fails care.data and the harm is ongoing to public trust and to confidence for researchers’ continued access to data.

The same failure was pointed out by the public members of the tiny Genomics England public engagement meeting two years ago in March 2014, called to respond to concerns over the lack of engagement and potential harm for existing research. The comms lead made a suggestion that the new model of the commercialisation of the human genome in England, to be embedded in the NHS by 2017 as standard clinical practice, was like steam trains in Victorian England opening up the country to new commercial markets. The analogy was felt by the lay attendees to be, and I quote, ‘ridiculous.’

Exploiting confidential personal data for public good must have support and good two-way engagement if it is to get that support, and what is said and agreed must be acted on to be trustworthy.

Policy makers must take into account broad public opinion, and that is unlikely to be submitted to a Parliamentary consultation. (Personally, I first knew such  processes existed only when care.data was brought before the Select Committee in 2014.) We already know what many in the public think about sharing their confidential data from the work with care.data and objections to third party access, to lack of consent. Just because some policy makers don’t like what was said, doesn’t make that public opinion any less valid.

We must bring to the table the public voice from past but recent public engagement work on administrative datasharing [5], the voice of the non-research community, and from those who are not stakeholders who will use the data but the ‘data subjects’, the public  whose data are to be used.

Policy Making must be built on Public Trust

Open policy making is not open just because it says it is. Who has been invited, participated, and how their views actually make a difference on content and implementation is what matters.

Adding controversial ideas at the last minute is terrible engagement, its makes the process less trustworthy and diminishes its legitimacy.

This last minute change suggests some datasharing will be dictated despite critical views in the policy making and without any public engagement. If so, we should ask policy makers on what mandate?

Democracy depends on social legitimacy. Once you lose public trust, it is not easy to restore.

Can new datasharing laws win social legitimacy, public trust and support without public engagement?

In my next post I’ll post look at some of the public engagement work done on datasharing to date, and think about ethics in how data are applied.

*************

References:

[1] The Royal Statistical Society data trust deficit

[2] “The social licence for research: why care.data ran into trouble,” by Carter et al.

[3] FAQs: Campaign for safe and ethical National Pupil Data

[4] CJEU Bara 2015 Ruling – fair processing between public bodies

[5] Public Dialogues using Administrative data (ESRC / ADRN)

img credit: flickr.com/photos/internetarchivebookimages/

Destination smart-cities: design, desire and democracy (Part four)

Who is using all this Big Data? What decisions are being made on the back of it that we never see?

In the everyday and press it often seems that the general public does not understand data, and can easily be told things which we misinterpret.

There are tools in social media influencing public discussions and leading conversations in a different direction from that it had taken, and they operate without regulation.

It is perhaps meaningful that pro-reform Wellington School last week opted out of some of the greatest uses of Big Data sharing in the UK. League tables. Citing their failures. Deciding they werein fact, a key driver for poor educational practice.”

Most often we cannot tell from the data provided what we are told those Big Data should be telling us. And we can’t tell if the data are accurate, genuine and reliable.

Yet big companies are making big money selling the dream that Big Data is the key to decision making. Cumulatively through lack of skills to spot inaccuracy, and inability to do necessary interpretation, we’re being misled by what we find in Big Data.

Being misled is devastating for public trust, as the botched beginnings of care.data found in 2014. Trust has come to be understood as vital for future based on datasharing. Public involvement in how we are used in Big Data in the future, needs to include how our data are used in order to trust they are used well. And interpreting those data well is vital. Those lessons of the past and present must be learned, and not forgotten.

It’s time to invest some time in thinking about safeguarding trust in the future, in the unknown, and the unseen.

We need to be told which private companies like Cinven and FFT have copies of datasets like HES, the entire 62m national hospital records, or the NPD, our entire schools database population of 20 million, or even just its current cohort of 8+ million.

If the public is to trust the government and public bodies to use our data well, we need to know exactly how those data are used today and all these future plans that others have for our personal data.

When we talk about public bodies sharing data they hold for administrative purposes, do we know which private companies this may mean in reality?

The UK government has big plans for big data sharing, sharing across all public bodies, some tailored for individual interventions.

While there are interesting opportunities for public benefit from at-scale systems, the public benefit is at risk not only from lack of trust in how systems gather data and use them, but that interoperability gets lost in market competition.

Openness and transparency can be absent in public-private partnerships until things go wrong. Given the scale of smart-cities, we must have more than hope that data management and security will not be one of those things.

But how will we know if new plans design well, or not?

Who exactly holds and manages those data and where is the oversight of how they are being used?

Using Big Data to be predictive and personal

How do we definde “best use of data” in “public services” right across the board in a world in which boundaries between private and public in the provision of services have become increasingly blurred?

UK researchers and police are already analysing big data for predictive factors at postcode level for those at risk or harm, for example in combining health and education data.

What has grown across the Atlantic is now spreading here. When I lived there I could already see some of what is deeply flawed.

When your system has been as racist in its policing and equity of punishment as institutionally systemic as it is in the US, years of cumulative data bias translates into ‘heat lists’ and means “communities of color will be systematically penalized by any risk assessment tool that uses criminal history as a legitimate criterion.”

How can we ensure British policing does not pursue flawed predictive policies and methodologies, without seeing them?

What transparency have our use of predictive prisons and justice data?

What oversight will the planned new increase in use of satellite tags, and biometrics access in prisons have?

What policies can we have in place to hold data-driven decision-making processes accountable?<

What tools do we need to seek redress for decisions made using flawed algorithms that are apparently indisputable?

Is government truly committed to being open and talking about how far the nudge unit work is incorporated into any government predictive data use? If not, why not?

There is a need for a broad debate on the direction of big data and predictive technology and whether the public understands and wants it.If we don’t understand, it’s time someone explained it.

If I can’t opt out of O2 picking up my travel data ad infinitum on the Tube, I will opt out of their business model and try to find a less invasive provider. If I can’t opt out of EE picking up my personal data as I move around Hyde park, it won’t be them.

Most people just want to be left alone and their space is personal.

A public consultation on smart-technology, and its growth into public space and effect on privacy could be insightful.

Feed me Seymour?

With the encroachment of integrated smart technology over our cities – our roads, our parking, our shopping, our parks, our classrooms, our TV and our entertainment, even our children’s toys – surveillance and sharing information from systems we cannot see  start defining what others may view, or decide about us, behind the scenes in everything we do.

As it expands city wide, it will be watched closely if data are to be open for public benefit, but not invade privacy if “The data stored in this infrastructure won’t be confidential.”

If the destination of digital in all parts of our lives is smart-cities then we have to collectively decide, what do we want, what do we design, and how do we keep it democratic?

What price is our freedom to decide how far its growth should reach into public space and private lives?

The cost of smart cities to individuals and the public is not what it costs in investment made by private conglomerates.

Already the cost of smart technology is privacy inside our homes, our finances, and autonomy of decision making.

Facebook and social media may run algorithms we never see that influence our mood or decision making. Influencing that decision making is significant enough when it’s done through advertising encouraging us to decide which sausages to buy for your kids tea.

It is even more significant when you’re talking about influencing voting.

Who influences most voters wins an election. If we can’t see the technology behind the influence, have we also lost sight of how democracy is decided? The power behind the mechanics of the cogs of Whitehall may weaken inexplicably as computer driven decision from the tech companies’ hidden tools takes hold.

What opportunity and risk to “every part of government” does ever expanding digital bring?

The design and development of smart technology that makes decisions for us and about us, lies in in the hands of large private corporations, not government.

The means the public-interest values that could be built by design and their protection and oversight are currently outside our control.

There is no disincentive for companies that have taken private information that is none of their business, and quite literally, made it their business to not want to collect ever more data about us. It is outside our control.

We must plan by-design for the values we hope for, for ethics, to be embedded in systems, in policies, embedded in public planning and oversight of service provision by all providers. And that the a fair framework of values is used when giving permission to private providers who operate in public spaces.

We must plan for transparency and interoperability.

We must plan by-design for the safe use of data that does not choke creativity and innovation but both protects and champions privacy as a fundamental building block of trust for these new relationships between providers of private and public services, private and public things, in private and public space.

If “digital is changing how we deliver every part of government,” and we want to “harness the best of digital and technology, and the best use of data to improve public services right across the board” then we must see integration in the planning of policy and its application.

Across the board “the best use of data” must truly value privacy, and enable us to keep our autonomy as individuals.

Without this, the cost of smart cities growing unchecked, will be an ever growing transfer of power to the funders behind corporations and campaign politics.

The ultimate price of this loss of privacy, will be democracy itself.

****

This is the conclusion to a four part set of thoughts: On smart technology and data from the Sprint16 session (part one). I thought about this more in depth on “Smart systems and Public Services” here (part two), and the design and development of smart technology making “The Best Use of Data” here looking at today in a UK company case study (part three) and this part four, “The Best Use of Data” used in predictions and the Future.

Breaking up is hard to do. Restructuring education in England.

This Valentine’s I was thinking about the restructuring of education in England and its wide ranging effects. It’s all about the break up.

The US EdTech market is very keen to break into the UK, and our front door is open.

We have adopted the model of Teach First partnered with Teach America, while some worry we do not ask “What is education for?

Now we hear the next chair of Oftsed is to be sought from the US, someone who is renowned as “the scourge of the unions.”

Should we wonder how long until the management of schools themselves is US-sourced?

The education system in England has been broken up in recent years into manageable parcels  – for private organisations, schools within schools, charity arms of commercial companies, and multi-school chains to take over – in effect, recent governments have made reforms that have dismantled state education as I knew it.

Just as the future vision of education outlined in the 2005 Direct Democracy co-authored by Michael Gove said, “The first thing to do is to make existing state schools genuinely independent of the state.”

Free schools touted as giving parents the ultimate in choice, are in effect another way to nod approval to the outsourcing of the state, into private hands, and into big chains. Despite seeing the model fail spectacularly abroad, the government seems set on the same here.

Academies, the route that finagles private corporations into running public-education is the preferred model, says Mr Cameron. While there are no plans to force schools to become academies, the legislation currently in ping-pong under the theme of coasting schools enables just that. The Secretary of State can impose academisation. Albeit only on Ofsted labeled ‘failing’ schools.

What fails appears sometimes to be a school that staff and parents cannot understand as anything less than good, but small. While small can be what parents want, small pupil-teacher ratios, mean higher pupil-per teacher costs. But the direction of growth is towards ‘big’ is better’.

“There are now 87 primary schools with more than 800 pupils, up from 77 in 2014 and 58 in 2013. The number of infants in classes above the limit of 30 pupils has increased again – with 100,800 pupils in these over-sized classes, an increase of 8% compared with 2014.” [BBC]

All this restructuring creates costs about which the Department wants to be less than transparent.  And has lost track of.

If only we could see that these new structures raised standards?  But,” while some chains have clearly raised attainment, others achieve worse outcomes creating huge disparities within the academy sector.”

If not delivering better results for children, then what is the goal?

A Valentine’s view of Public Service Delivery: the Big Break up

Breaking up the State system, once perhaps unthinkable is possible through the creation of ‘acceptable’ public-private partnerships (as opposed to outright privatisation per se). Schools become academies through a range of providers and different pathways, at least to start with, and as they fail, the most successful become the market leaders in an oligopoly. Ultimately perhaps, this could become a near monopoly. Delivering ‘better’. Perhaps a new model, a new beginning, a new provider offering salvation from the flood of ‘failing’ schools coming to the State’s rescue.

In order to achieve this entry to the market by outsiders, you must first remove conditions seen as restrictive, giving more ‘freedom’ to providers; to cut corners make efficiency savings on things like food standards, required curriculum, and numbers of staff, or their pay.

And what if, as a result, staff leave, or are hard to recruit?

Convincing people that “tech” and “digital” will deliver cash savings and teach required skills through educational machine learning is key if staff costs are to be reduced, which in times of austerity and if all else has been cut, is the only budget left to slash.

Self-taught systems’ providers are convincing in their arguments that tech is the solution.

Sadly I remember when a similar thing was tried on paper. My first year of GCSE maths aged 13-14  was ‘taught’ at our secondary comp by working through booklets in a series that we self-selected from the workbench in the classroom. Then we picked up the master marking-copy once done. Many of the boys didn’t need long to work out the first step was an unnecessary waste of time. The teacher had no role in the classroom. We were bored to bits. By the final week at end of the year they sellotaped the teacher to his chair.

I kid you not.

Teachers are so much more than knowledge transfer tools, and yet by some today seem to be considered replaceable by technology.

The US is ahead of us in this model, which has grown hand-in-hand with commercialism in schools. Many parents are unhappy.

So is the DfE setting us up for future heartbreak if it wants us to go down the US route of more MOOCs, more tech, and less funding and fewer staff? Where’s the cost benefit risk analysis and transparency?

We risk losing the best of what is human from the classroom, if we will remove the values they model and inspire. Unions and teachers and educationalists are I am sure, more than aware of all these cumulative changes. However the wider public seems little engaged.

For anyone ‘in education’ these changes will all be self-evident and their balance of risks and benefits a matter of experience, and political persuasion. As a parent I’ve only come to understand these changes, through researching how our pupils’ personal and school data have been commercialised,  given away from the National Pupil Database without our consent, since legislation changed in 2013; and the Higher Education student and staff data sold.

Will more legislative change be needed to keep our private data accessible in public services operating in an increasingly privately-run delivery model? And who will oversee that?

The Education Market is sometimes referred to as ‘The Wild West’. Is it getting a sheriff?

The news that the next chair of Oftsed is to be sought from the US did set alarm bells ringing for some in the press, who fear US standards and US-led organisations in British schools.

“The scourge of unions” means not supportive of staff-based power and in health our junior doctors have clocked exactly what breaking their ‘union’ bargaining power is all about.  So who is driving all this change in education today?

Some ed providers might be seen as profiting individuals from the State break up. Some were accused of ‘questionable practices‘. Oversight has been lacking others said. Margaret Hodge in 2014 was reported to have said: “It is just wrong to hand money to a company in which you have a financial interest if you are a trustee.”

I wonder if she has an opinion on a lead non-executive board member at the Department for Education also being the director of one of the biggest school chains? Or the ex Minister now employed by the same chain? Or that his campaign was funded by the same Director?  Why this register of interests is not transparent is a wonder.

It could appear to an outsider that the private-public revolving door is well oiled with sweetheart deals.

Are the reforms begun by Mr Gove simply to be executed until their end goal, whatever that may be, through Nikky Morgan or she driving her own new policies?

If Ofsted were  to become US-experience led, will the Wild West be tamed or US providers invited to join the action, reshaping a new frontier? What is the end game?

Breaking up is not hard to do, but in whose best interest is it?

We need only look to health to see the similar pattern.

The structures are freed up, and boundaries opened up (if you make the other criteria) in the name of ‘choice’. The organisational barriers to break up are removed in the name of ‘direct accountability’. And enabling plans through more ‘business intelligence’ gathered from data sharing, well, those plans abound.

Done well, new efficient systems and structures might bring public benefits, the right technology can certainly bring great things, but have we first understood what made the old less efficient if indeed it was and where are those baselines to look back on?

Where is the transparency of the end goal and what’s the price the Department is prepared to pay in order to reach it?

Is reform in education, transparent in its ideology and how its success is being measured if not by improved attainment?

The results of change can also be damaging. In health we see failing systems and staff shortages and their knock-on effects into patient care. In schools, these failures damage children’s start in life, it’s not just a ‘system’.

Can we assess if and how these reforms are changing the right things for the right reasons? Where is the transparency of what problems we are trying to solve, to assess what solutions work?

How is change impact for good and bad being measured, with what values embedded, with what oversight, and with whose best interests at its heart?

2005’s Direct Democracy could be read as a blueprint for co-author Mr Gove’s education reforms less than a decade later.

Debate over the restructuring of education and its marketisation seems to have bypassed most of us in the public, in a way health has not.

Underperformance as measured by new and often hard to discern criteria, means takeover at unprecedented pace.

And what does this mean for our most vulnerable children? SEN children are not required to be offered places by academies. The 2005 plans co-authored by Mr Gove also included: “killing the government’s inclusion policy stone dead,” without an alternative.

Is this the direction of travel our teachers and society supports?

What happens when breakups happen and relationship goals fail?

Who picks up the pieces? I fear the state is paying heavily for the break up deals, investing heavily in new relationships, and yet will pay again for failure. And so will our teaching staff, and children.

While Mr Hunt is taking all the heat right now, for his part in writing Direct Democracy and its proposals to privatise health – set against the current health reforms and restructuring of junior doctors contracts – we should perhaps also look to Mr Gove co-author, and ask to better understand the current impact of his recent education reforms, compare them with what he proposed in 2005, and prepare for the expected outcomes of change before it happens (see p74).

One outcome was that failure was to be encouraged in this new system, and Sweden held up as an exemplary model:

“Liberating state schools would also allow the all-important freedom to fail.”

As Anita Kettunen, principal of JB Akersberga in Sweden reportedly said when the free schools chain funded by a private equity firm failed:

“if you’re going to have a system where you have a market, you have to be ready for this.”

Breaking up can be hard to do. Failure hurts. Are we ready for this?
******

 

Abbreviated on Feb 18th.

 

The front door to our children’s personal data in schools

“EdTech UK will be a pro-active organisation building and accelerating a vibrant education and learning technology sector and leading new developments with our founding partners. It will also be a front door to government, educators, companies and investors from Britain and globally.”

Ian Fordham, CEO, EdTech UK

This front door is a gateway to access our children’s personal data and through it some companies are coming into our schools and homes and taking our data without asking.  And with that, our children lose control over their safeguarded digital identity. Forever.

Companies are all “committed to customer privacy” in those privacy policies which exist at all. However, typically this means they also share your information with ‘our affiliates, our licensors, our agents, our distributors and our suppliers’ and their circles are wide and often in perpetuity. Many simply don’t have a published policy.

Where do they store any data produced in the web session? Who may access it and use it for what purposes? Or how may they use the personal data associated with staff signing up with payment details?

According to research from London & Partners, championed by Boris Johnson, Martha Lane-Fox and others in EdTech, education is one of the fastest growing tech sectors in Britain and is worth £45bn globally; a number set to reach a staggering £129bn by 2020. And perhaps the EdTech diagrams in US dollars shows where the UK plan to draw companies from. If you build it, they will come.

The enthusiasm that some US EdTech type entrepreneurs I have met or listened to speak, is akin to religious fervour. Such is their drive for tech however, that they appear to forget that education is all about the child. Individual children. Not cohorts, or workforces. And even when they do it can be sincerely said, but lacks substance when you examine policies in practice.

How is the DfE measuring the cost and benefit of tech and its applications in education?

Is anyone willing to say not all tech is good tech, not every application is a wise application? Because every child is unique, not every app is one size fits all?

My 7-yo got so caught up in the game and in the mastery of the app their class was prescribed for homework in the past, that she couldn’t master the maths and harmed her confidence. (Imagine something like this, clicking on the two correct sheep with numbers stamped on them, that together add up to 12, for example, before they fall off and die.)

She has no problem with maths. Nor doing sums under pressure. She told me happily today she’d come joint second in a speed tables test. That particular app style simply doesn’t suit her.

I wonder if other children and parents find the same and if so, how would we know if these apps do more harm than good?

Nearly 300,000 young people in Britain have an anxiety disorder according to the Royal College of Psychiatrists. Feeling watched all the time on-and offline is unlikely to make anxiety any better.

How can the public and parents know that edTech which comes into the home with their children, is behaviourally sound?

How can the public and parents know that edTech which affects their children, is ethically sound in both security and application?

Where is the measured realism in the providers’ and policy makers fervour when both seek to marketise edTech and our personal data for the good of the economy, and ‘in the public interest’.

Just because we can, does not always mean we should. Simply because data linkage is feasible, even if it brings public benefit, cannot point blank mean it will always be in our best interest.

In whose best Interest is it anyway?

Right now, I’m not convinced that the digital policies at the heart of the Department for Education, the EdTech drivers or many providers have our children’s best interests at heart at all. It’s all about the economy; when talking if at all about children using the technology, many talk only of ‘preparing the workforce’.

Are children and parents asked to consent at individual level to the terms and conditions of the company and told what data will be extracted from the school systems about their child? Or do schools simply sign up their children and parents en masse, seeing it as part of their homework management system?

How much ‘real’ personal data they use varies. Some use only pseudo-IDs assigned by the teacher. Others log, store and share everything they do assigned to their ID or real email address , store performance over time and provide personalised reports of results.

Teachers and schools have a vital role to play in understanding data ethics and privacy to get this right and speaking to many, it doesn’t seem something they feel well equipped to do. Parents aren’t always asked. But should schools not always have to ask before giving data to a commercial third party or when not in an ’emergency’ situation?

I love tech. My children love making lego robots move with code. Or driving drones with bananas. Or animation. Technology offers opportunity for application in and outside schools for children that are fascinating, and worthy, and of benefit.

If however all parents are to protect children’s digital identity for future, and to be able to hand over any control and integrity over their personal data to them as adults,  we must better accommodate children’s data privacy in this 2016 gold rush for EdTech.

Pupils and parents need to be assured their software is both educationally and ethically sound.  Who defines those standards?

Who is in charge of Driving, Miss Morgan?

Microsoft’s vice-president of worldwide education, recently opened the BETT exhibition and praised teachers for using technology to achieve amazing things in the classroom, and urged innovators to  “join hands as a global community in driving this change”.

While there is a case to say no exposure to technology in today’s teaching would be neglectful, there is a stronger duty to ensure exposure to technology is positive and inclusive, not harmful.

Who regulates that?

We are on the edge of an explosion of tech and children’s personal data ‘sharing’ with third parties in education.

Where is its oversight?

The community of parents and children are at real risk of being completely left out these decisions, and exploited.

The upcoming “safeguarding” policies online are a joke if the DfE tells us loudly to safeguard children’s identity out front, and quietly gives their personal data away for cash round the back.

The front door to our children’s data “for government, educators, companies and investors from Britain and globally” is wide open.

Behind the scenes  in pupil data privacy, it’s a bit of a mess. And these policy makers and providers forgot to ask first,  if they could come in.

If we build it, would you come?

My question now is, if we could build something better on pupil data privacy AND better data use, what would it look like?

Could we build an assessment model of the collection, use and release of data in schools that could benefit pupils and parents, AND educational establishments and providers?

This could be a step towards future-proofing public trust which will be vital for companies who want a foot-in-the door of EdTech. Design an ethical framework for digital decision making and a practical data model for use in Education.

Educationally and ethically sound.

If together providers, policy makers, schools at group Trust level, could meet with Data Protection and Privacy civil society experts to shape a tool kit of how to assess privacy impact, to ensure safeguarding and freedoms, enable safe data flow and help design cybersecurity that works for them and protects children’s privacy that is lacking today, designing for tomorrow, would you come?

Which door will we choose?

*******

image credit: @ Ben Buschfeld Wikipedia

*added February 13th: Oftsed Chair sought from US