Top Ten Problems with UK Information Sharing Proposals

Chris Pounder of Amberhawk information consultants sends me his Top Ten Problems with the British government´s new information-sharing proposals that are to be found buried deep in the Coroners and Justice Bill, where perhaps they thought no-one would notice… these are part of much lengthier and more thorough analysis submitted to the Joint parliamentary Committee on Human Rights (JCHR), which explains why the proposals ignore or conflict with the recommendations of 2008´s Data Sharing Review conducted by Richard Thomas and Mark Walport for the Ministry of Justice itself. These are sumarised by me here, and any errors and omissions are therefore my own:

  1. Lack of scrutiny. There is no provision for the JCHR to scrutinise this (or any other) wide-ranging statutory power which impacts on Article 8 of the European Convention on Human Rights (ECHR), nor any attempt to explain how this provision is consistent with human rights legislation.
  2. The extension of information sharing beyond personal data. The use of “any person” in the Bill means that it applies to information sharing by any public or private body or individual. “Information sharing” powers are not limited to personal data and the person who receives the shared information might be a foreign government or organisation. [for example the FBI´s proposed Server in the Sky]
  3. The “exceptional” may become the routine The Data Sharing Review recommended that the sharing of personal data should be legitimised in exceptional circumstances. However, in the Bill there is instead a legitimation of general information sharing, whenever it falls within a “relevant policy objective” [which is basically anything a Minister decides].
  4. The generality of an Information Sharing Order. There is no limit as to how “person”, “purpose” and “information class” are specified in an Order. There is no explicit requirement for the purpose of the information sharing to be one of those specified in Article 8(2) ECHR.
  5. The prospect of unlimited data sharing from large Government databases. The Bill appears to facilitate data sharing from any Government database without Parliament being explicitly informed of this sharing when an Order is before Parliament. The prohibition in the clause only relates to Part 1 of the Regulation of Investigatory Powers Act (RIPA). By implication, sharing from other national databases (e.g. the national identity register of the ID Cards Act) does not need to be explicitly mentioned in an Order. This means that unlimited data can be shared from these other national databases by means of a general order-making provision.
  6. The exclusion of critical comment on the purpose of the processing. In the Bill, the Information Commissioner is not allowed to comment on whether “the sharing of information enabled by the order is necessary to secure a relevant policy objective”. The effect is to inhibit the Commissioner from commenting on the purpose of the processing, which is the main purpose of the Information Commissioner! Plus, because this applies to more than personal information, much of the proposed sharing is outside his remit.”
  7. The range of the powers. The powers are widely drawn and their application is very broad. There is no explicit provision in the main sharing provisions which would facilitate data subject rights and freedoms (e.g. right to object ; need to obtain consent). Instead, these provisions can “modify” the application of any law (including the Data Protection Act and the Human Rights Act) which will weaken the protection afforded to data subjects.
  8. The lack of transparency. There is no obligation to disclose to the Information Commissioner or Parliament any background document or legal advice about a proposed Information Sharing Order. There is no obligation to answer any formal request for information from the Commissioner. There is no obligation to engage the public on the subject of a draft Information Sharing Order.
  9. The irrelevance of the proposed Code of Practice. There is nothing in these information sharing clauses which expressly states that the sharing of personal data has to be consistent with the proposed non-statutory Code of Practice. The Code is not subject to approval by Parliament; rather, it is subject to approval by the Secretary of State (SoS).There is no provision which sets out what happens if there is a disagreement between SoS and Information Commissioner about the content of a Code. There is no active role for Parliament in relation to the content of a Code.
  10. Orders can be implemented to achieve purely administrative objectives. For example, suppose Ministers are told by civil servants that the problems associated with one of the Government’s big database projects would be resolved if they used criminal convictions from the Police National Computer. The Bill allows the Minister to argue that the sharing was necessary to secure a policy objective, it was proportionate as there was no other way of securing the policy objective (abandoning a large IT project is not an option), and it was in the public interest to secure the policy objective (given the amount of money committed to the project). This means that sharing which could be excessive and disproportionate in terms of Article 8 becomes necessary and proportionate in terms of realising a policy objective.

Previously, I commented that No2ID were overstating their case that this proposal was the greatest threat to information rights after the ID Register. After reading Chris´s analysis, I think they might be underestimating its importance. The creation of a generalised and weakly accountable ability for the state to share information of any kind with any one they wish, is a far greater threat than the creation of any single database, however extensive. I disagree with their views on the Data Sharing Review, but No2ID’s data sharing site still has the best summary of proposals and action people can take…

Brazil as Surveillance Society? (1) Bolsa Família

The claim that Brazil is a surveillance society, or at least uses surveillance in the same fundamental organising way as the UK or Japan does, is based on the bureaucracy of identification around entitlement and taxation, rather than policing and security.

My previous post on the subject of whether Brazil was a surveillance society put one side of an argument I am having with myself and colleagues here: that the use surveillance in Brazil is fundamentally based on individual (and indeed commodified and largely class-based) security, rather than surveillance as fundamental social organising principle (as one might legitimately claim is the case in Britain). Now, I deliberately overstated my case and, even as I was posting, my argument was being contradicted by colleagues in the same room!

So here´s the counter-argument – or at least a significant adjustment to the argument. In most nation-states, entering into a relationship with the state involves forms of surveillance by the state of the person. This relationship is more or less voluntary depending on the state and on the subject of the relationship. In most advanced liberal democracies, the nature of surveillance is based on the nature of citizenship, particularly:

  1. the ability of citizens to establish claims to entitlement, the most fundamental to most being a recourse to the law (to protect person and property), secondly the ability to case a vote, and more something that is generally more recent in most states, the right to some kind of support from the state (educational, medical, or financial);
  2. the ability of the state to acquire funds from citizens through direct or indirect taxation, to support the entitlements of citizens, and to maintain order.

I am not going to consider law and order, or indeed electoral systems, here but rather I will concentrate on the way that surveillance operates in an area I had previously begun to consider: the bureaucracy of identification around state-citizen relations particularly in the areas of entitlement and taxation. The claim that Brazil is a surveillance society, or at least uses surveillance in the same fundamental organising way as the UK or Japan does, is based on this rather than policing and security.

There are two broad aspects: on the one side, taxation, and on the other, entitlement. I´ll deal first with the latter (which I know less about at the moment), in particular in the form of Lula´s Programa Bolsa Família (PBF, or Family Grant Program), one of the cornerstones of the socially progressive politics of the current Brazilian government. The PBF provides a very simple, small but direct payment to families with children, for each child, provided that the children go to school and have medical check-ups.

Of course these requirements in themselves involve forms of surveillance, through the monitoring of school attendance by children – for which there is a particular sub-program of the PBF called Projeto Presença (Project Presence) with its own reporting systems – and epidemiology and surveillance of nutrition through the Ministério de Saúde (Ministry of Health). However underlying the entitlement is massive compulsory collection of personal information through the Cadastro Único para Programas Sociais (CadÚnico, or Single Register for Social Programs), set up by Lula´s first administration to unify the previous multiple, often contradictory and difficult to administer number of social programs. This is, of course a database system, which as the CadÚnico website states, ¨funciona como um instrumento de identificação e caracterização socioeconômica das famílias brasileiras¨ (¨functions as an means of identification and socioeconomic caracterization of Brazilian families¨). Like most Brazilian state financial systems, CadÚnico is operated through the federal bank, the Caixa Econômica Federal (CAIXA). The CadÚnico database is founded on ¨um número de identificação social (NIS) de caráter único, pessoal e intransferível¨ (¨a unique, personal and non-transferable Social Identification Number or NIS¨). I am unclear yet how this NIS will relate to the new unique identification system for all citizens.

The PBF Card
The PBF Card

Entitlement is demonstrated with (yet another!) card, the patriotic yellow and green Cartão PBF. Like the CPF card, this is a magnetic strip card rather than a smart card, and is required for all transactions involving the PBF. Also like the CPF, but unlike many other forms of Brazilian ID, it has nothing more than the name of the recipient and the CadÚnico number printed on it. In this case the recipient is generally the mother of the children being claimed for, a progressive and practical measure shared with other family entitlement programs in Brazil.

Happy smiling PBF cardholders!
Happy smiling PBF cardholders!

The PBF card in itself may not be enough to claim as you would still need at least the Registro Geral (national ID) card to prove that you are the named holder of the PBF card. The card itself may be simply designed to generate a sense of inclusion, as the pictures of happy smiling PBF cardholders on the government websites show consistently emphasise, although of course, like so many other markers of entitlement to state support, it could also become a stigma.

The information collection to prove entitlement is quite extensive, and here I have translated roughly from the website:

  • house characteristics (number of rooms; construction type; water, sewerage and garbage systems);
  • family composition (number of members, dependents like children, the elderly, those with physical handicaps);
  • identification and civil documents of each family member;
  • educational qualification of each family member;
  • professional qualifications and employment situation;
  • income; and
  • family outgoing (rent, transport, food and others).

Although PBF is a Federal program, the information is collected at the level of individual municipalities, and there is thus the potential for errors, differences in collection methods, delays and so on to hamper the correct distribution of the money. So each municipality is required to have a committee called the Instância de Controle Social (Social Control Authority) which, whilst it may sound sinister to anglophone ears, actually refers to the control of civil society over the way that the government carries out its social programs. This is also quite a lot of information of the most personal kind and whilst, unlike in many countries there is no central authority of Commissioner for Data Protection in Brazil, there is particularly for PDF, an Observatório de Boas Práticas na Gestão do Programa Bolsa Família (Observatory for Best Practice in the Management of the PBF), which has a whole raft of measures to safeguard and protect the data, correct errors etc (what has been called habeus data principles). Effectively, this is a case of knowing exactly quis custodis ipsos custodes!

Now of course, such a large database of information about the most vulnerable people in society has the potential to be misused by a less progressive or even fascist government. Marxist analysis of early welfare systems has tended to colour our views of such programs as being solely about the management of labour on behalf of capital and the control of the working classes by the state to prevent them from more revolutionary action. For more recent times in Surveillance Studies, John Gilliom´s book, Overseers of the Poor, showed how much Federal assistance programs in the USA could impact negatively upon the lives of claimants, particularly women, in the Appalachian region, and revealed the everyday forms of resistance and adaptation that such women used to make the programs function better for them. I will have to examine more detailed anthropological studies of the PBF to see whether similar things are true of the Brazilian program. I don´t want to get too much into the effectiveness of this program now, although I am trying to examine the correlation of the PBF with apparently declining crime rates in Brazilian cities, but it is worth noting that the World Bank rates it as one of the most successful ways of dealing with extreme poverty in the world. As a general observation, it does seem that only those who object to redistributive policies full stop (or just dislike Lula himself) or those who think it does not go far enough, have any serious complaint about the PBF. But there is far more to consider here…

Identity and Identification in Brazil (continued)

…the Brazilian driving licence is a goldmine of personal information…

I spent a little while over the last couple of days examining the actual material identity documents currently required in Brazil. Here are some pictures with a little explanation. There will be a lot more in the final article!

The first is the simplest but in many ways the most important to life-chances. This is the Cadastro de Pessoas Físicas (CPF) (Register of Physical (or Natural) Persons) card (or Taxpayer’s Card).


‘Pessoas Físicas’ is a a piece of legalese that is draws a distinction between humans and other ‘legal persons’, like corporations or governments. The CPF number is issued to all those who pay tax and is essential if one wants any formal work. The actual document is a blue plastic card like old-style credit cards, which also has a machine readable magnetic strip on the back.

The number is also required for many other government transactions, and it is, apparently a major disaster if you lose the card, or if for some reason, your CPF number is rescinded (which can happen if you don’t pay tax in Brazil for more than a year, for example if you are abroad, without explanation). Many people who live in the favelas, and who are involved in the shadow economy do not have a CPF, which is a severe obstacle to social inclusion.

The second document is the Registro Geral (General Registry) (ID) card, a double-sided piece of thick paper, just larger than a credit card. It is oriented vertically at the front and horizontally at the back.

RG card

As I noted in the first post I made on this subject, the RG card cross-references the CPF and also birth certification (it lists the full names of both mother and father and city and state of origin). This card is the one that is being replaced by the new RIC smartcard ID system.

Finally, we have the Carteira Nacional de Habilitação, the driving licence which, despite its name, is issued at state rather than national-level. The colour and format differs from state-to-state, however they all have pretty much the same level of information (a lot!) and cross-identification with other forms of ID. This one is from Paraná, which is a paper usually folded in half horizontally. It is specifically forbidden to laminate it.


The Brazilian driving licence is a goldmine of personal information. Partly this is because the licence had been intended to be a unifying piece of identification (a practice typical of ‘autocentric’ cultures!), containing all the information on both the CPF card and the RG card, and more. This will now not be the case following the issuing of the new RIC cards, so it will be interesting to see if the quantity of information on these licences will be reduced or, if not, what the justification will be for having this much visible personal information on one paper document.

The new Brazilian ID system

The new Brazilian ID-card
The new Brazilian ID-card (from Renato Siqueira's Conversa Digital)

There are more details of the new Brazilian ID card and system on Renato Siqueira’s Conversa Digital blog, including some informative images and photos. It seems that far from eliminating the various different numbers currently used, this new system will merely create a kind of overlay. And, not only that, but the CPF, RG and electoral number will be printed on the back. Unless every single transaction will actually require the taking of fingerprints or the verification of photos, this card will be even more of a convenient source of personal information to thieves and fraudsters than ever before. Plus the chip technology is the same standard format that has proved to easy to clone and access illicitly elsewhere…

Identity and Identification in Brazil

My host and colleague here at PUCPR, Rodrigo Firmino, and I are working on a small bit of research and a paper for The Second Multidisciplinary Workshop on Identity in the Information Society (IDIS 09), at the the London School of Economics, on June 9th this year.

Our paper is based around a case of identity theft, which is endemic in Brazil, which we use to open up the laws, practices and technologies of identification here. One thing that is already clear is that Brazil is a highly bureaucratic state – for example, the forms you need to fill in just to get a mobile phone are incredible in their detail – yet the forms of identification which one needs for every transaction with the state and many private organisations too, are highly insecure.

One example is that every personal cheque has printed on it not only the usual information (bank name and address, bank sort code, account holder name and account number), but also has the 11-digit Cadastro de Pessoas Fisicas (CPF) (a taxpayer’s card) number and the 9-digit Registro Geral (RG) (the national ID card) number. This must be a utter joy to fraudsters and identity thieves!

What’s more, all these are not just numbers in a database somewhere but physical documents in their own right, and on each there is a lot of this cross-identification: the CPF card also has the name and date of birth, the CPF number is ubiquitous, appearing also on the RG card and the driving licence. The latter has its own 11-digit registration number, but also has the RG number, name, and place and date of birth. What is even more interesting is that the RG card not only contains a photo and a thumbprint (the state database contains prints of all 10 fingers and thumbs), but also the names of both parents. This means it can be related more easily to the birth certificate. It reminds me a little of the Japanese system which still prioritises the family above the individual in some ways, but there is no actual equivalent of the koseki, the Japanese family register.

Now, in the name of security and “para integrar os bancos de dados de diversos órgãos dos sistemas de identificação do Brasil” (to inegrate the databases of the diverse organisations of identification systems in Brazil), the Ministry of Justice is proposing to merge some of these – the RG, CPF, Driving Licence and Electoria Regisirtation, into a new, smart, Registro de Identidade Civil (RIC) card based on a unique number. Whilst this will have many of the same problems as new smart ID systems everywhere else, at the very least it might stop Brazilian citizens carrying around multiple documents that list almost everything thieves and fraudsters need and can access without any sophisticated equipment. The process is due to start now, and run until 2017, so we will be taking a look at this as it proceeds.

I’ll put some pictures up with explanations later today…