The Co-Ordinated Online Record of Electors (CORE) - the Implementation of National Access Arrangements

The Co-ordinated Online Record of Electors (CORE) - The implementation of national access arrangements

Consultation Paper CP 29/05 14/12/2005 This consultation will end on 07/03/2006

The Co-ordinated Online Record of Electors (CORE) The implementation of national access arrangements

A consultation produced by the Department for Constitutional Affairs. This information is also available on the DCA website at www.dca.gov.uk

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Contents

Executive summary 7 Introduction 10 The proposals 12 Appendix A: Extract from feasibility & options report 52 Appendix B: Diagram showng prospective relationship of CORE to other systems 68 About you 76 How to respond 77 The Consultation Criteria 79 Consultation Co-ordinator contact details 80

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Executive summary

Part 1 of the Electoral Administration Bill1 currently before Parliament makes provision for the establishment of one or more Co-ordinated Online Record of Elector (CORE) schemes. It will allow - at national level - access to electoral registration data that will continue to be locally gathered and maintained. The scheme would be established by a secondary legislation order, a draft of which would first need to be actively approved by both Houses of Parliament. This consultation paper seeks views on what the CORE scheme order should contain. If the CORE provisions in the Bill obtain Royal Assent, we would wish to bring forward the first CORE scheme order – drafted in the light of responses to this paper - as soon as possible after that. That is why we are consulting on these issues now.

Through the establishment of a CORE scheme, the Government is seeking to achieve: improvements in the integrity of electoral registers; more efficient access to registration data for those authorised; support modernisation of the voting process; improvements in the efficiency of the electoral administration process; and provide a mechanism by which reports can be provided and/or research conducted efficiently on electoral registration data, based on a national dataset. The establishment of a CORE scheme also aims to meet recent recommendations made by both the independent Electoral Commission and the OSCE’s2 Office for Democratic Institutions and Human Rights.

• Narrowing down the options (page 14)

Six basic approaches to building CORE have been identified. By assessing each against the criteria of functionality, likely acceptability, implementability, and risk, the Government is proposing a structure that retains independent local systems, but requires them to mirror their information to a central data repository.

• Compiling the record (page 19)

1 http://www.publications.parliament.uk/pa/cm200506/cmbills/050/2006050.htm

2 Organisation for Security and Cooperation in Europe

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

The paper then proposes that the types of information that should pass between local electoral registration officers (EROs) and the CORE ‘keeper’ should be not just information from the ‘full’ version of the electoral register itself, but also absent voter list information, and any additional electoral registration information that the Electoral Administration Bill is currently proposing for EROs (i.e. personal identifiers and ‘off-register’ details for proposed ‘anonymous electors’). We also suggest that marked register information might also form part of the CORE record, recognising that this would be a longer term aim.

Also in relation to the flow of information between local ERO and CORE keeper, we propose that the ability to use the Election Markup Language data transfer standard should be made mandatory upon EROs by the end of 2006, and that CORE should primarily link to EROs using the Government Connect arrangements. Finally in this section, we seek views on the proposals that: a) EROs should send largely automated updating information to the CORE keeper on a daily basis; and b) that such information should be accepted into the central record automatically, so that as complete a record as possible is available at all times.

• Providing copies of the record from CORE (page 30)

The paper seeks views on whether – and to what extent – a CORE keeper may be subject to different constraints than those that apply to a local ERO. Specifically, we propose that large-scale users of electoral data should obtain that information from the CORE keeper rather than local EROs, and we seek views on whether smaller-scale users should also be required to do so. We do not, at this time, propose that the data held by a CORE keeper should be directly provided to a returning officer to run an election.

We propose that - under strict controls - authorised bodies (e.g. the police) should be given the ability to directly access the CORE record electronically in order to search for the most up to date information. We invite views on who might be appropriate bodies to allow such access to.

• Elector access to CORE (page 36)

We propose to allow a householder to confirm electronically via a CORE keeper that no change is required to the electoral register information held for a property, with that information then being forwarded to the relevant ERO. We do not propose to allow householders to add, amend, or delete information in the same way. However, we do propose that – subject to appropriate security - an individual

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

should be able to check, confirm, and request changes to their information that a CORE keeper held, with the CORE keeper again subsequently forwarding confirmations and change requests to the relevant EROs.

• Anti-fraud mechanisms (page 39)

Using the different types of information proposed to be passed to CORE by local EROs (see above) we believe that the CORE keeper should be empowered to check his records for apparently anomalous data relating to: a) the same elector being registered in more than one place; b) the appointment of the same individual to act as proxy for more than two people; c) the same address being used as the mailing address for the postal voting packs of more than a specified number of electors; and d) the same elector being issued with and/or casting more ballots than they are entitled to do in a given election.

We then propose that the keeper should be required to forward all relevant details of anomalous data to all EROs who it affects, and that EROs be required to act in consequence of – and respond to – any such notification from a CORE keeper. We invite views on how frequently such notifications should be sent out by a CORE keeper, and on what the consequent impact of such a requirement to act would be on EROs. We believe it should also be possible for an ERO to be pro-active in initiating a duplicate check with a CORE keeper on a specific elector.

• Establishing the first CORE scheme (page 45)

The paper then looks at how the first CORE scheme might be established and seeks views on: a) whether the CORE infrastructure should be established on a limited geographical scale initially, or be introduced across the UK from the start; and b) whether the Electoral Commission or some other public body should be appointed as the CORE ‘keeper’.

• Future possibilities (page 48)

Finally, in relation to longer term possibilities, we seek views on whether linkage between the CORE dataset and other public sector databases should be actively pursued. The potential for receiving updates on changes to the status of individual electors and undertaking data matching could significantly improve both the integrity and comprehensiveness of electoral registers, but equally presents fairly significant data protection and technical issues that would need to be resolved.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Introduction

This paper sets out for consultation a number of detailed issues and proposals for the implementation of arrangements providing national access to locally collected and maintained electoral registration data. The consultation is aimed at those involved in producing electoral register data (electoral registration officials, the providers of their software, and the electors who supply the information to the register in the first place) and the users of that data (e.g. government departments, the Electoral Commission, political parties, registered credit reference agencies), as well as other agencies and organisations throughout the United Kingdom that have an interest in the proposed changes.

This consultation is being conducted in line with the Code of Practice on Consultation issued by the Cabinet Office and falls within the scope of the Code. The Consultation Criteria, which are set out on page 79 have been followed.

A final Regulatory Impact Assessment which considered all parts of the Electoral Administration Bill (including Part 1 on the Coordinated Online Record of Electors (CORE) was published on 11 October 2005. This is available on the DCA website at: http://www.dca.gov.uk/legist/electroadminbill.pdf. A detailed RIA on CORE will be published when the secondary legislation required to implement it is consulted on. Consequently, this paper does not contain a Partial Regulatory Impact Assessment. If you disagree with this conclusion you are invited to send your reasons as part of your overall response to this paper.

Copies of the consultation paper are being sent to:

Organisation for Security and Cooperation in Europe (OSCE) Office for Democratic Institutions and Human Rights

Disability Rights Commission

Electoral Commission

Association of Electoral Administrators

Scottish Assessors Association

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Society of Local Authority Chief Executives and Senior Managers

Society of Local Authority Lawyers and Administrators in Scotland

Local Government Association

Political Parties

Members of the House of Commons Constitutional Affairs Select Committee

Credit reference agencies (e.g. Experian, Equifax, CallCredit)

Voluntary organisations (e.g. SCOPE, Operation Black Vote, Hansard Society, Liberty)

Police Information Technology Organisation

Scottish Executive

Welsh Assembly Government

Office for National Statistics

General Register Office for Scotland

However, this list is not meant to be exhaustive or exclusive and responses are welcomed from anyone with an interest in or views on the subject covered by this paper.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

The proposals

Background

1. The Co-ordinated Online Record of Electors (CORE) project was launched in 2004. Its aim is to: a) standardise key elements of the different types of software used to maintain electoral registers by some 400 locally-based electoral registration officers across the UK (Phase 1); and b) building on standardisation, to establish arrangements to provide access at a national level to authorised persons and organisations. From a single central point, individual electors should be able to access, confirm, and potentially request changes to their own electoral registration details online, while authorised users should be able to obtain copies of the data. (Phase 2).

2. The ultimate objectives of CORE are:

• help improve the integrity of electoral registers;

• give convenient national access to standardised electoral registration data that continues to be locally collected and maintained;

• help support modernisation of the voting process;

• increase the efficiency of electoral administration;

• provide for the preparation of reports centrally, based on the national dataset.

3. Additionally, the CORE project aims to give effect to certain recommendations from:

• the Electoral Commission’s May 2003 report ‘The electoral registration process’, that “electoral registers should continue to be compiled and managed locally, but should form part of a national register” and “electronic, online registration be enabled"3; and

3 http://www.electoralcommission.org.uk/files/dms/Registration_9799-7973__E__N__S__W__.pdf (pages 19-20)

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

• the August 2005 OSCE Office for Democratic Institutions and Human Rights’ report on the UK’s General election of 2005 that “a state-wide database for registered voters could be a useful tool to identify or prevent possible multiple registrations.” 4

4. A consultation paper primarily on local standardisation (Phase 1) issues was published in May 20045. A summary of responses and statement of how the Government is moving forward in consequence will be published shortly.

5. The May 2004 consultation paper also sought general views on national access (Phase 2). Those responses have helped to identify further issues and to inform the drafting of this paper, which sets out more detailed policy proposals on CORE implementation. We welcome general views on whether these proposals are right in principle and practice and, if not, what alternatives should be considered. We have included a number of specific questions, on which we would welcome responses. A list of the questions and details of how to respond are set out at pages 71-77 of this document.

4 http://www.osce.org/documents/odihr/2005/08/15921_en.pdf (page 5)

5 Co-ordinated Online Register of Electors (CORE) – Standardising Electoral Registration. http://www.odpm.gov.uk/index.asp?id=1138050

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Narrowing down the options

6. In autumn 2004 the Government commissioned Accenture e-Democracy Services to conduct a detailed feasibility and options study into the approach that should be taken to deliver Phase 2. Accenture were able to call on their own direct experience of delivering state-wide voter register systems in the USA in response to the requirements of the Help America Vote Act (‘HAVA’) - passed following the Presidential elections of 2000.

7. Delivered in late February 2005, the Accenture report6 looked at the contextual background in the UK and the project’s requirements and identified a range of six possible models. These ranged from a highly localized model to a highly centralized model. Appendix A is an extract from the Accenture report that sets out the process of differentiating the six models and describes how each works. In summary:

• Model 1 – Registration Message Broker. Each ERO’s system sends data that may be relevant to one or more other EROs to a central point. The central point then forwards the data to the relevant EROs for their consideration and action. No data is retained centrally, but there is nevertheless likely to be a high volume of transactions.

• Model 2 – Indexed Registration Search Engine. Very similar to Model 1, but the central point has a data index that it uses to more accurately direct messages between EROs. The index can be refined and updated continually by the messages themselves. Electoral registration data as such is not held in the centre.

• Model 3 – Central Warehouse Distributed Transaction. Essentially Model 1 or 2, but with a ‘data warehouse’ added on. This data warehouse is the first instance of proper electoral registration data being held at the central point. However, the warehouse is only populated (and subsequently updated) by the content of the transactions as they go through the central point, and there is no mechanism to check the integrity of the data.

6 Key findings of the report are reproduced here, but copies of the report itself can be made available on request to the same contact specified for responses to this consultation paper. Accenture were asked to include in their final report detailed information on how they - as possible contractors - would go about delivering the models we favoured (in order to establish a reasonable estimate of what implementation costs would be). As Accenture’s implementation process is considered commercially confidential, this information is edited from the publicly available version of the final report.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

• Model 4 – Central Integrity Reporting Repository. A complete record of all electoral registration data is held from the start in a central dataset and is kept current by updating from local ERO systems. The central infrastructure also includes an ‘integrity engine’ then checks the data it is sent against specified rules, to identify possible anomalous records.

• Model 5 – Central Registry. Similar to Model 4, but instead of merely ‘flagging up’ potential anomalous data for investigation by local EROs, this model would actively enforce integrity rules. That is, the system would not allow an apparently anomalous entry into the central record.

• Model 6 – Central System. The most centralised model in that this structure does not allow local EROs to use their own systems locally. Instead, there is merely one database run from a central location. Local EROs would manage ‘their’ segment of the data on the central database through ‘dumb’ terminals that were part of that central system.

8. The six models were then each assessed against four separate key feasibility criteria:

• Functionality – to what extent can a solution deliver the functionality required to meet CORE objectives?

• Acceptability – how acceptable is each solution likely to be with respect to legislation and stakeholder concerns? That is, to what extent is additional legislation likely to be unnecessary, and how amenable would most stakeholder groups probably be to each model? It should be noted that there are a range of stakeholders on this topic, with quite different – and sometimes directly conflicting – points of view. As a numerically based ‘average view’ the table below therefore reflects primarily the view of the numerically largest group of those whose views Accenture sought - EROs and their staff7.

• Implementability – how easy would it be to bring sufficient technical capability and capacity to bear to implement the solution?

7 To demonstrate the difficulty in assessing this category in particular, the trend in acceptability was reversed for national users of electoral registration data, due to their concern that more localised models would be far less likely to deliver the information they need in an effective and efficient manner – one of the key objectives of the CORE project and the original founding objective of its predecessor (LASER).

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

• Risk – how risky is each solution, accounting for risks relating to technological failure, security failure, operational failure and failure of a solution to adapt to change?

9. The following table summarises the outcome of this assessment process, using the key - L (Low), M (Medium), H (High):

Model Functionality Acceptability Implementability Risk

1) Message Broker L H M M-H

2) Indexed Search L H L M-H Engine

3) Central Data L-M M L-M H Warehouse

4) Central Integrity H M H L-M Reporting Repository

5) Central Registry H L M-H L

6) Central System H L L-M L

10. The list of options was then narrowed down to the two that seemed most desirable in light of the criteria above. The four models rejected and the reasons why were:

• Model 1: Very low functionality, and medium-high risk. In particular, this model delivers little – if anything – in terms of improved register integrity, convenient national access for data users, or improved efficiency for administrators.

• Model 2: Only a slight improvement in terms of functionality for those key benefits sought from CORE, with still a medium-high risk and even lower implementability than model 1.

• Model 3: Better functionality, but still much worse than more centralised models, implementability still relatively low, and the highest risk of any of the models.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

• Model 6: A relatively low implementability score compared to other centralised models, but this model’s critical failure against the criteria was the extent to which it removed control over the register from local EROs.

11. By contrast, both Models 4 and 5 provided a genuinely centralised co-ordinated record of electors (and therefore scored highly in delivering the anticipated functionality requirements), and also achieved high implementability and low risk scores. We therefore identified these as our preferred models, with Model 4 ranked higher than Model 5, as the latter removes from local EROs an element of control over the data and could therefore be viewed as conflicting with that aspect of the CORE requirements.

Q1: Do you agree that Models 4 and 5 are best suited to meet the aims and objectives of Phase 2 of the CORE project? If you disagree, please state which model you believe would be a better fit and set out your reasons.

12. Subsequently, Accenture provided a detailed assessment of how Models 4 or 5 could be delivered, including the technical hardware, software and people requirements, the estimated implementation time, and the estimated cost of delivery. It was specifically recognised that a competitor could well take a different approach, with a consequential effect on the cost in time and resources. However, we felt that it was important to at least establish reliable initial estimates before progressing Phase 2 any further substantively.

13. Flowing from the Accenture report and using the information it contains, we have further refined the detail of the proposals for how CORE Phase 2 might be implemented. As we will see from the rest of this paper, the preferred model on which we are now consulting is in essence a Central Integrity Reporting Repository (Model 4). Where relevant in the rest of this paper, we have explained why we have preferred the approach of that model over that of the Central Registry (Model 5).

14. In addition to those principled policy reasons for favouring Model 4 over Model 5, cost is likely to be a serious issue were we to pursue Model 5. In their report, Accenture provided estimated overall implementation costs, using both a ‘High estimate’ and ‘Low estimate’ figure for Model 4 and 5. For full implementation, Model 4 estimates were £7.4M at the low end and £15.1M at the high end, whilst Model 5 was £11.1M at the low end and £22M at the high end. CORE currently has a budget for implementation of national access arrangements of circa £10M, so it will be seen that whilst Model 4 may present some issues at the high end of costs, it should be well within budget at the low end. By 17

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

contrast, even at the low end of costs, Model 5 would exceed the current CORE budget and at the likely high end could cost more than double the funds currently available.

15. We believe primary legislation would be required to allow any genuine national access arrangements to be established. The general principles and parameters of how a ‘CORE scheme’ - and those involved with it - should operate are specified in the current Electoral Administration Bill itself. We propose that affirmative secondary legislation (a ‘CORE scheme order’) should then set out the operational detail of a particular scheme.

16. This approach would allow the detail of how a CORE scheme operates to be established or – if appropriate - amended with relatively little demand on the Parliamentary timetable. Just as importantly, it ensures that any such secondary legislation would be drawn up in the light of responses to this consultation and would be actively debated and approved by both Houses of Parliament.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Compiling the record

17. The Government’s preferred model is one that establishes a comprehensive and up-to-date consolidated record of the national electoral registration dataset, bringing together in one place copies of the information held locally by all the EROs within the area covered by the CORE scheme. In this section we are seeking views on two practical questions:

• What types of electoral registration information should be held in a CORE scheme?; and

• How should electoral registration data be transmitted to a CORE scheme?

What types of information?

18. The first question to ask about a CORE scheme is what information should it hold. Fundamentally, should CORE hold only information that would be found on the electoral register itself, or should it also hold other electoral information that EROs are responsible for, such as the lists of absent voter preferences?

• Register information

19. Clearly, any consolidated record of electors will need to hold basic information about the electors themselves currently required on an electoral register. We propose this should be all of the information that you would expect to find on the ‘full’ - rather than ‘edited’ – register.

20. Accordingly, for the area it covers, CORE would hold for all registered electors their name, address (if in the UK), polling district, and elector number. Where relevant, information would also be held on an elector about: the date of their 18th birthday (for attainers); any jury-exempt status due to age (does not apply in Scotland); their desire to ‘opt-out’ from inclusion of their details in the ‘edited’ register (indicated by a ‘Z’ marker); or any restricted franchise status (e.g. for citizens of most other EU states, for overseas electors, or for members of the House of Lords).

21. As with the register of electors itself, information about individual electors should also be linked to the polling district within which their voting entitlement arises, and the electoral areas that each polling district forms a part of, e.g. the ward/division, local authority, constituency (whether for elections to the UK

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Parliament or other bodies), parish/community council area, and National Park area in Scotland that that polling district forms part of.

22. We would propose that all of the information that appears in the full register of electors should be replicated on the CORE system. Given the range of users that would be seeking to obtain information from CORE in the future, who would otherwise be getting copies of the register from local EROs, we see no reason to remove certain types of data from the record before it is transmitted to CORE. Additionally, as much identifying information as possible will be necessary to help ensure that the same individual is being identified in any duplicate-matching process (see separate later chapter dealing with ‘Anti-fraud mechanisms’).

23. Clearly, where a data user is only entitled to have a copy of the edited register, CORE would need to filter out those electors who had indicated their wish to ‘opt-out’ of the edited register. It is for this reason that we propose CORE must receive the opt-out marker information from local registers.

Q2: Do you agree that all information that currently appears on the full register of electors should be held in the CORE? If you disagree, please state which information from the full register you believe should NOT be held in CORE, and set out your reasons.

• Marked register information

24. The marked register is essentially a copy of the full register of electors (as above) that has been marked to show that an elector has been issued with a ballot paper (either at a polling station or by post). In certain circumstances, such marked register information will denote that an individual has actually voted, e.g. from any future marked register of postal votes returned, or ‘marked registers’ that could be produced by future electronic voting systems. It is important to note that the marked register shows only that an elector has had their voting paper(s). It does not show who the elector voted for.

25. Currently, the responsibility for producing the marked register lies with a returning officer, and a specified public official is responsible for retaining it for a specified period after the election to which it relates. However, the Government is currently proposing that responsibility for retention on the marked register should in future lie with EROs across the UK (except in Scotland). Accordingly, a record of such marked register information might also be usefully transmitted by EROs to a consolidated CORE dataset in future. If

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

this could be done, integrity checking could again be used to identify where the same individual had appeared to vote more than they were entitled to in any given election.

26. We recognise that at the current time marked registers are generally only kept in hard copy form, i.e. they are a print-out of the register that is marked in manuscript at the time that ballot papers are issued. Accordingly, there would need to be a fundamental change of approach if the information was to be imported into a CORE system. As this is likely to involve significant further work and resources to achieve, it is unlikely that any inclusion of marked register information in a CORE system could realistically be deliverable short-term.

27. However, we are nevertheless keen in principle to gain the benefit of an integrity check against fraudulent duplicate voting. We would therefore welcome views – particularly from electoral administrators and their software suppliers - on what the likely additional cost to them might be to get marked register information into electronic form, and whether we should therefore look to meet such costs in order to gain the benefit to the integrity of elections that would result.

Q3: If responsibility for retention of the marked register is in future to be the responsibility of EROs, what is your estimate of the cost of getting such information imported to a consolidated CORE system.Do you think good value for money andbenefits to the integrity of the election would result?

• Absent voter list information

28. Lists kept under Schedule 4 to the Representation of the People Act 2000, contain information as to ‘absent voting’ arrangements for electors. The information consists of:

• the identity of electors who have chosen to vote by post and the address – if different from their ‘qualifying address’ that appears in the register itself – to which their postal voting papers are to be sent;

• the identity of electors who have appointed a proxy and the name and address of that proxy; and

• the identity of any such proxies who have chosen to vote by post and their mailing address.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

29. These lists also contain information about the time periods – or types of elections – for which the absent voting rights are to be applicable, e.g. for a particular election or type of election, and for a defined or indefinite period.

30. Again there would be potential benefit in having this kind of information consolidated. For postal voting details, comparison of mailing addresses from postal voting lists (items a) and c) above) would reveal where the same address was being used to send multiple postal voting packs (which may be an attempt at postal voting fraud). For proxy voting, no individual may act as proxy for more than two electors (unless they are immediate family members), so a comparison could be made in a consolidated dataset to identify any potential infringement of this rule.

Q4: Do you agree that information set out in the statutory absent voter lists maintained by EROs should be supplied to a consolidated CORE system? If you disagree, please set out your reasons.

• Additional elector information.

31. The Government is currently proposing that additional information be held on electors separate to that held on the ‘main register’ and absent voter lists. Specifically, it is proposed that additional ‘personal identifiers’ could be collected to help prevent and counter possible electoral fraud (e.g. signature and/or date of birth for all electors). At the time of writing, the Electoral Administration Bill proposes only that there be piloting of the collection of personal identifiers in the registration process, although there is also provision for it to be rolled out, if appropriate following successful piloting.

32. The collection and use of these additional personal identifiers is a matter on which there is ongoing discussion. On the one hand, it is recognised that additional personal identifiers may help improve the security of the electoral process and may help counter any attempted fraud. However, set against this is the fact that a requirement to provide additional information decreases the likelihood of an elector participating in the registration process at all, as it makes the task of completing the necessary forms more complicated, particularly if every individual has to complete their own separate form. There is therefore a balance to be struck between security and participation and the Government feels that it is right to test out different approaches before determining where the correct balance lies.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

33. If such personal identifier information is to be collected in future, we would propose that it should also be communicated to and held on a CORE system. We feel this is particularly important, as personal identifiers help ensure that when conducting any internal data searches to find matching records for individuals (see section below dealing with anti-fraud mechanisms), the CORE system can do so with a much greater degree of accuracy than is currently possible. Indeed, without such personal identifiers, much of the prospective benefit of CORE for improving the integrity of electoral registers is likely to be unworkable in practice.

34. For example, a search to detect all occurrences of ‘John Smith’ would be a fairly pointless task, given the likely number of matches that would be identified and need further investigation. However, a list matching only John Smiths born on a particular date would be a much shorter list, which would be much more realistic to investigate. Similarly, if a signature is to be collected from each elector on registration, this too could be scanned into a local system (most registration software packages already provide this facility) and the image transmitted to CORE. It could then be used as an additional check to ensure that an apparent multiple occurrence of an individual was indeed the same person.

35. Set against this, if the view was taken that the proposed anti-fraud functionalities of CORE (see later section) were not worth pursuing, then it is difficult to see why any additional personal identifiers should be provided to a central CORE system.

36. If future additional personal identifiers were supplied to CORE, we believe policy on whether a CORE keeper could make such information available to third parties should follow the policy on whether an ERO could make the information available.

Q5: Do you agree that any additional personal identifier information on individuals that is gathered during electoral registration should be supplied to a consolidated CORE system? If you disagree, please set out your reasons. Views on what types of personal identifier might be particularly useful and realistic for use in a CORE context would also be welcomed.

37. It is also proposed that a special category of ‘anonymous elector’ be created, whose personal details would not appear in the main register, but would be held separately. We would also propose that these details should be communicated to and held by CORE. If they were not, an individual registered 23

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

as an anonymous elector would be able to avoid all of the integrity checking that we are proposing to apply to all other electors in the CORE system.

38. If an individual registers anonymously it suggests that they want as few people as possible to know their details. Accordingly, we would propose that the CORE system should be subject to the same strict rules about restricting access to this personal data as are intended to apply to the originating ERO.

Q6: Do you agree that any future anonymous elector details should be supplied to a consolidated CORE system, subject to the same restrictions on access as apply to the originating ERO? If you disagree, please set out your reasons.

Transmitting the data

• Election Markup Language

39. The mechanism by which all of this information would need to be transmitted from local ERO systems to the central CORE system would be the Election Markup Language (EML) data transfer standard, which is based on the increasingly common XML framework. In essence, EML establishes a standard structure for data presentation when sending a routine electoral message electronically from one system to another.

40. The standard structure allows any recipient system to know what structures to expect, and such a system can therefore be designed to allow much greater automation – and therefore efficiency – to be introduced to an electoral process: the system can recognise the purpose of a data transmission from its structure, and need only then follow a pre-programmed routine that tells it what to do with the information contained in a transmission of that type. As noted above, much work has already gone into Phase 1 of the CORE project, which has been seeking to introduce use of the EML standard into all systems used by EROs in the UK.

41. At the time of writing, the vast majority of EROs were expected to have EML- compliant software by the end of 2005, in line with the timetable we have established for CORE Phase 1. One or two others have indicated that they will become EML-compliant during the first half of 2006. The use of the single data transfer standard by all is fundamental to Phase 2, as a network of systems using different software (as proposed) will not otherwise work efficiently. Whilst

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

we would normally prefer to rely on voluntary take-up, the fundamental importance of 100% compliance in this context we believe justifies making the ability to use EML mandatory upon every ERO in the UK.

42. A mechanism already exists to enable the creation of such a duty on EROs, under the direction-making power of section 52 (1) and (1A) of the Representation of the People Act 1983. Given that progress with Phase 1 of CORE (see above) is already well advanced, we would therefore propose to issue a direction under these powers, to make the use of EML-compliant software by EROs mandatory by the end of 2006. Whilst this is likely to be well before introduction of a national access infrastructure, it would give time for EROs to adapt to the routine use of EML.

Q7: Do you agree that the Government should make use of EML-compliant software mandatory on EROs by the end of 2006 to enable complete UK coverage of CORE Phase 1 electoral registration software standardisation? If you disagree, please set out your reasons.

• Linking EROs and a central CORE system

43. One issue which has yet to be resolved in relation to the transmission of data from local EROs to CORE is the data transport and related security requirements that should be applied to it. Our initial view is that we should seek to utilise appropriate existing public sector networks wherever possible, rather than seek to construct from scratch a separate network, due to the additional cost that would otherwise be required.

44. Relatively recently, a programme of work known as ‘Government Connect’ has been established, which may well prove to be a highly suitable infrastructure that CORE data exchange could utilise. The detail of the Government Connect programme is set out in its Prospectus, which can be viewed and downloaded from the Government Connect website8, or a hard copy obtained from the Government Connect programme office9. In broad terms, the part of the aims and objectives of the Government Connect programme that seems to offer greatest potential to the implementation of the CORE project is that focused on “supporting joint working and communication”, by:

8 www.govconnect.gov.uk

9 Telephone 0845 838 2945

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

• “enabling closer working through secure communication;

• joining up back-office systems; and

• consolidating reporting to central government.”

45. Specifically, the ‘GC Exchange’ aspect of the Government Connect programme seems to offer a suitable network mechanism for linking up the electoral registration systems of local authorities with each other and with a central CORE system for routine data transfer. Additionally, the ‘GC Mail’ aspect could be useful for any more ad hoc communication that may be needed between the operators of the different systems.

46. As Government Connect is a relatively new initiative, there is a question on whether it will be in place in time for CORE to benefit from it. The Government Connect prospectus sets a target date of December 2007 for all local authorities to be using Government Connect. In practice it is not likely that a CORE system will be operational before that date, given the need to secure primary and secondary legislation, procure a contractor and build the system and train the users before going live.

47. The Government Connect project is still developing in terms of geographical coverage and the CORE vision ultimately requires all EROs in the UK to be linked into the same network. Accordingly, whilst we would propose to actively pursue the possibility of using the Government Connect network for CORE data transactions, we do not at this stage propose to close down the possibility of exploring other options including, for example, the small-scale trialling of alternatives.

Q8: Do you agree that the Government should actively pursue the possibility of using the Government Connect network for CORE data transactions, whilst also – for the time being - exploring the viability of alternative networking approaches? If you disagree, please set out your reasons and what approach to establishing a suitable network you would prefer.

• Frequency of updates

48. How up to date should CORE data be? The Government’s general approach to CORE has been that it should follow – insofar as possible – the generic rules that apply to the electoral registration process, without introducing new or different requirements unless there is specific justification.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

49. The current rules for electoral registration allow electors to apply to go on the register at any time, either during the annual canvass period or – outside of that time – via an individual rolling registration form. Applications may be processed on a continuous basis by EROs – including entering the details on their local database software. However the entry onto the register does not legally come into effect until the next appropriate ‘monthly update’ of the register is published, or until the 1st December annual update if the application is received during the autumn canvass period. There is also a statutory five-day period to allow for any objections to be made to an application for registration before the information even enters this monthly update procedure. This may suggest that there would be no need for EROs to update the CORE system with any more frequency than for normal publication of statutory monthly and annual updates.

50. However, there may be many users of electoral registration data who would prefer that more frequent updating of a CORE system should take place. In part this may be because whilst updates to the register itself only legally take effect on a monthly basis, changes to other records maintained by the ERO and potentially reflected on CORE take immediate effect, e.g. changes to the lists of absent voting entitlement.

51. Additionally, data users may wish to have the most up-to-date information about people going onto the register even if they are – in technical electoral terms - not yet actually ‘on the register’. For instance, the ERO for an area that an elector was moving out of would wish to be notified of this departure before the elector is fully registered in the area that they have moved to. This would be done to ensure that the elector is not improperly registered twice: CORE could only send such a notification if it first received information from the ERO in the elector’s new area. The notification would have to clearly state that the elector was only at that time ‘provisionally’ registered pending the next monthly update. We would welcome views on whether information should be sent to CORE by local EROs on applications for registration as soon as they have been received, i.e. before the five-day objections period has expired and before the ERO has formally made a ‘determination’ that the person should be added to the next monthly register.

52. At this time the Government is minded to propose that an ERO should be required to send an update to CORE each working day, even if that update merely confirms that there has been no change since the previous day’s notification. We do not view this as a move to ‘continuous registration’, as updating information would need to specify the monthly update at which an

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

individual’s details were to be formally added to the register. We recognise that this may appear to put more of a burden on EROs, but as it should be possible to make such standard data collection and transmission processes almost wholly automated, a minimal additional burden is anticipated in comparison with the benefits generated.

53. The main alternative would seem to be a requirement to update CORE on a monthly basis. Whilst this would not seem to require any additional effort on the part of an ERO (especially if CORE is to handle subsequent supply of monthly updates to other recipients of the data), it does mean that CORE could be working with out of date information. This could hinder its effective and meaningful duplication checking and provision of accurate information to electors seeking to check their details online (see sections on ‘Elector access’ and ‘Anti-fraud mechanisms’ below).

54. Please express more simply:A daily update therefore seems to us to be a reasonable requirement, striking the right balance between a genuinely ‘real- time’ – i.e. instant - updating requirement and allowing the CORE record to become inappropriately out of step with the local ERO’s information.

Q9: Do you agree that EROs should send updates to a CORE central system on a daily basis? If you disagree, please set out your reasons and what frequency of updating you would prefer to see instead.

• Acceptance into the CORE ‘live’ record

55. One of the key points that distinguishes the Central Integrity Reporting Repository (CIRR) and Central Registry (CR) models in the feasibility and options study is the level of control applied to data before it enters the ‘live’ record in the central dataset. In the CIRR model, so long as the data is properly formatted in line with the requirements of EML (see above), it will be fully accepted into the ‘live’ record. It is only subsequently that any potential ‘anomalies’ that CORE may detect – such as the apparent presence of the same elector on more than one register - would be notified to local EROs.

56. By contrast, the CR system uses a centrally determined set of rules to apply a ‘filter’ to data as it is received. If the data does not satisfy the rule – e.g. it appears to be information on an elector already in the ‘live’ record – the data is effectively put into ‘quarantine’ and not let into the ‘live’ record until it has been double-checked and confirmed.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

57. Although we are not currently proposing to do so (see paragraph 66 below), if the register to run an election was extracted from the central record of electors, the subtle difference of approach between the two models could become a key issue. This is because the elector whose details are in quarantine at the time the register is extracted for use at an election would potentially be disenfranchised, even if their data were subsequently confirmed as accurate.

58. Even though causing only a temporary potential delay, central rules to filter certain data from EROs appear to infringe on the principle that a local ERO should remain responsible for what does and does not get included in the register. If a quarantine approach was adopted, there could be a growing disparity between what is held on local electoral registers and what appears in the central record.

59. Consequently, the Government’s inclination is to adopt the CIRR approach and allow data straight into the live record, with integrity checking and reporting occurring subsequently. However, we would welcome views on whether, instead, data should be quarantined before being admitted to the record.

Q10: Do you agree that data sent by EROs should go straight into the CORE record, with subsequent integrity checking and reporting of possible anomalies? If you disagree, please set out your reasons.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Providing copies of the record from CORE

60. A key function of CORE will be to provide copies of the register and other information held by EROs to those entitled to receive it. As described above, this could include marked registers (to the extent they may in future be held by EROs post-election) and absent voter lists. Detailed regulations10 already exist on who is entitled to this information (in particular who can receive the ‘full’ version of the register rather than the edited version), what they can do with it, and whether a fee can be charged for its supply.

61. In broad terms, the Government proposes that these rules should apply to a CORE keeper11 in his/her dealings with the data supplied by an ERO in the same way that they apply to the ERO him/herself. However, there are some specific points on which we might suggest that CORE takes a slightly different approach and on which we welcome views. These are set out below. Additionally, where in the future data users may be able to purchase electoral registration information from CORE, we welcome views on what proportion of the income from such sales should go back to EROs who provided the information (on a pro rata basis). Our initial view is that 100% should be passed back, but we would be interested to hear whether there is any support for the CORE keeper retaining a percentage (please say how much) in respect of processing the sale, which could help offset the running costs of CORE itself.

Large-scale users of registration data

62. On the face of it, a CORE keeper would become an alternative to local EROs as a source of register information for users of such data. As a CORE will cover an area pertaining to a number of EROs, we might therefore assume that large- scale ‘national’ users of such data will probably prefer to go to CORE for the information in future, rather than approach EROs separately. Such ‘national’ users of the ‘full’ copy of the register would encompass bodies such as the

10 The Representation of the People (England and Wales) Regulations 2001 (as amended) and the equivalent Regulations made for Scotland and Northern Ireland.

11 A ‘keeper’ would be the person legally designated as being responsible for running the CORE system.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Electoral Commission, the headquarters of registered political parties, Government Departments (in particular the Office for National Statistics and General Register Office for Scotland), and registered credit reference agencies. There may also be other organisations interested in acquiring an ‘edited’ copy of the register for the whole area covered by the CORE scheme.

63. One of the possible benefits of CORE is that it could alleviate local EROs of the burden of dealing with numerous requests for copies of their register and thus free up some of their time for more ‘value-added’ registration duties. We therefore suggest that such ‘national’ users should in future only obtain copies from a CORE keeper where a CORE scheme is operational and no longer be able to approach the local ERO.

Q11: Do you agree that, in areas where a CORE scheme is operational, specified large-scale users of electoral registration data should no longer be able to obtain such data direct from local EROs? If you disagree, please set out your reasons.

Smaller scale users

64. There are also a number of smaller scale or local users of registration data. These are often not interested in registration data beyond the area covered by one or two EROs. Examples here would include local branches of political parties, independent election candidates, and - in relation to the ‘edited’ register - local commercial concerns. Some electoral administrators state that it is mainly these smaller – but more numerous – local users whose requests for copies of the register make the biggest demands on ERO resources.

65. In addition to the best allocation of the administrative burden of providing copies of the register, it is also worth considering the question of consistency of approach in interpreting and applying the rules surrounding provision of copies. For instance, since the rules were first introduced, there have been reports of uncertainty and different interpretations of the rules by different EROs, e.g. in determining whether or not an organisation is a credit reference agency and therefore entitled to a copy of the full register. In this regard, it may be more consistent for a single body to provide all copies of registers..

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

copies either from their local ERO or CORE; or c) only be able to obtain the copies from the local ERO? Please set out your reasons.

The returning officer

66. We do not propose that returning officers should receive the electoral register from CORE. Under the proposed model, even with – say – daily updating of the central record, the legal ‘register’ maintained by the local ERO and the ‘record’ of it held on CORE will possibly differ. Although such a discrepancy will invariably be only temporary (until the next update is sent through), if the elector information needed to actually run an election is taken from CORE rather than the local ERO, it could potentially disenfranchise electors whose details are in fact recorded on the local register. We therefore propose that elector information for running an election is taken from the local ERO rather than the CORE.

67. One solution would be to adopt a different basic model: either fully localised (Model 1, where there is no central information and access would be direct to the local registers) or fully centralised (Model 6, where there is only a central system and there are no separately-held local registers). However, the Government does not favour either of these models, for the reasons set out in the ‘Narrowing down the options’ section of this paper (add paragraph reference).

68. If ultimate responsibility for the data were to transfer in future to the CORE keeper from the local ERO, a CORE ‘top-down’ infrastructure would be required. This would be along the lines of Model 5 or 6 as set out in the Accenture report referred to in the ‘Narrowing down the options section above). It would make using the data from the central record of electors to actually run an election a much more viable prospect.

69. However, for as long as the local ERO remains the person ultimately responsible for the data in the register, the Government’s view is that the elector information used to run an election should continue to be provided directly from the ERO to the returning officer. It is important to note here that CORE is designed to improve this particular activity: The main objective of Phase 1 is to standardise the transfer of data. Implementation of Phase 2 should therefore not affect it.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

70. Potential future developments in remote electronic voting or polling stations allowing eligible electors to vote in any election even in a different electoral area might prove difficult without a central CORE election register system. In both cases, a consolidated register of electors will need to be accessible to the e-voting software and/or the polling station staff to confirm entitlement to vote.

71. However, this only becomes relevant where the election is conducted across areas covered by multiple EROs, so would not seem to be an issue for local elections. Furthermore, with larger scale elections, the problem is merely one of collating the data from the affected EROs at that time, rather than requiring a single source to be readily available via CORE. As such data should be produced in standard form anyway following CORE Phase 1, this should only be a question of ensuring that all EROs supply the required information to the returning officer at the required time, something the law already requires.

Q13: Do you agree with the proposal that returning officers should continue to obtain the register and other information required to conduct the election direct from the relevant EROs, rather than from CORE? If you disagree, please set out your reasons.

Inspection copy of the register

72. Current regulations require an ERO to make a copy of their ‘full’ register available for inspection (under supervision) in appropriate locations in the geographical area for which they are responsible. CORE is designed to provide primarily electronic data copies of register information. Whilst we would not preclude CORE from providing the relevant information in paper form instead, we are not inclined to require a CORE keeper to make a copy of all registers for general public inspection.

73. Such a requirement would place an additional administrative burden on a CORE keeper, as it would entail making suitable arrangements and supervisory staff available for on-site personal inspection by the public. The current provision is intended to allow an elector to visit a local office to ensure in person that they are properly registered and – if relevant – check that no other person is improperly registered. These needs seem best served by locally- based EROs, and so making an identical requirement for the more geographically remote central CORE keeper seems unnecessary.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Q14: Do you agree with the proposal that a CORE keeper should not be subject to the requirements for making a copy of the full register available for personal public inspection? If you disagree, please set out your reasons.

Direct access to CORE for authorised bodies

74. A number of data users are likely to express interest in the possibility of accessing directly ‘online’ the most up to date information on an individual held by CORE. For example, in conducting a credit check, a credit reference agency may wish to check an elector’s details with the most up-to-date information CORE.

75. Larger organisations are likely to want both direct access and bulk data, as they often use the latter to compile a historical record (e.g. a ‘credit history’ developed by credit reference agencies). However, some users may wish to have direct access only: for example a smaller registered political party making a small number of donation checks in accordance with the Political Parties Elections and Referendums Act 2000, or possibly the police investigating a particular individual. Accordingly, we are minded to establish a mechanism by which authorised bodies can directly access CORE for the purpose of conducting a check on an individual elector or address.

76. It is not proposed that any such organisation should be given access to information on CORE that they would not otherwise be entitled to receive under current arrangements. Access to the full register by named authorised individuals within authorised organisations would be very tightly controlled, Each user may need to undergo some form of security clearance in advance and have their own unique log-in details.

77. We would welcome views on how such a browse or search facility should work for external users connecting directly to CORE. As controls on data from the full register and its uses would apply, we would propose that a browse facility and/or search by name or other personal identifier would be appropriate. In our view this would accord with current regulations barring the copy of the full register made available for general public inspection from being searchable electronically by name: a search initiated by an individual not entitled to access full register data would only access information on the ‘edited’ register.

78. For users who would normally be expected to pay for copies of the register, a charging structure would need to be established for online access. We would

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

welcome views on how such charging arrangements should operate, e.g. whether there should be a general licence fee, or ‘charge per access/copy’, and what the level of charges should be.

Q15: Do you agree with the proposal that authorised bodies should be granted direct electronic access to the CORE central dataset to browse and/or initiate an electronic search for an individual record? If yes, we would welcome your views on how a charging structure might work for those who are normally expected to pay for copies of the register. If you disagree, please set out your reasons.

79. We would also be especially interested in the views of respondents on who exactly should and should not be given such direct access, particularly in comparison with those currently entitled to receive a copy of the full register of electors. For example, our initial view is that organisational users with an ongoing need to access electoral register information should have direct access. This would include: the Electoral Commission, registered political parties, registered credit reference agencies, and the police.

80. We believe returning officers should also be given direct access, perhaps on an ‘as needed’ basis tied to the length of the election period. However, we do recognise that the benefits of this will be determined by the size of the electorate. For example, the returning officer at a local election would probably gain most from going directly to a local ERO, while the Greater London Returning Officer for Greater London Authority elections and a Regional Returning Officer at European Parliamentary elections would find a central record more could be of significant administrative benefit. to

81. Other individuals might be entitled to a copy of the full public register. However it may not be appropriate to provide them with direct access to a CORE system, for example, if they are election candidates. It is likely that the individual would prefer to get their copy of the register from the local ERO, so it is not clear what benefit they would gain from direct access to CORE. Conversely, the more individuals and organisations that are authorised for direct access to CORE, the more problematic it becomes for the CORE scheme to administer that access and bringing a potentially increased risk of inappropriate access being granted.

Q16: If direct access to a CORE system for the purpose of confirming identity were to be established, we welcome views on who exactly should be given the ability to use that facility. 35

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Elector access to CORE

82. One particularly useful potential function of CORE is to act as a central point through which electors in any area covered by the CORE scheme can interact to some degree with their local electoral register.

Householder registration

83. Currently, a number of EROs use online registration facilities during the annual canvass period. Unique log-in details are printed onto the canvass form. This form is pre-completed with the most recent information held on electors at each address. If no change to those details is required then a member of the household can go online (or telephone) and use the log-in details to confirm this. This information is then passed to the relevant ERO. However, it is not currently legal to amend or add details by this procedure.

84. We propose that CORE should replicate this service, so that a single site and/or telephone number would cover all the EROs in the CORE area, Currently this service is provided by various separate commercial providers selected by individual EROs. This would generate additional work for a CORE keeper, but as this would amount to forwarding notice of confirmations to relevant EROs it should be possible to largely automate the process. A benefit of centralising the process is that an elector moving from one area to another would not need to familiarise themselves with a different registration process.A single system with potentially UK-wide coverage could be used in a large-scale promotional campaign to encourage voting.

85. We do not propose to allow a householder to amend or add details to a canvass form via an online or telephone service. T

Q17: Do you agree with the proposal that online/telephone access to CORE should be made available for householders to confirm the accuracy of a pre- completed canvass form (for forwarding to the relevant ERO), but not to make changes (including adding or deleting electors)? If you disagree, please set out your reasons.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Rolling or individual registration

86. We are also considering an online equivalent of the rolling registration process. We propose that an individual elector should be able to log into CORE and be able to see all the information that CORE – and therefore the local ERO - holds on them. This would include not just the main register information, but also any absent voting preferences, marked register information (i.e. whether or not a ballot paper was issued to them and – possibly – a postal vote received back ), and any ‘off-register’ information such as anonymous elector details and personal identifiers that CORE may hold. We propose that an elector would then be able to confirm the accuracy of any of this information or – if appropriate – request a correction.

87. Subject to appropriate security being in place, the Information Commissioner’s Office has been supportive of the proposal for an individual to access the information held about them online for the purpose of checking and either confirming accuracy or requesting changes. Additionally, it is worth noting that countries with comparable electoral arrangements already provide an online facility for individual electors to check their own details12.

88. It may be that consultees would prefer a two-stage approach. In the first stage individual electors could be given read-only access to view the information and confirm accuracy, but with requests to change the information handled through a paper-based system. A form could be completed online with CORE, but would still need to be printed out, signed, and sent in hard copy to the ERO. The second stage could be to make the requests for change a fully online process. Although our inclination is to implement both aspects at the same time, we would welcome views on this two-stage approach.

89. We propose that confirmations or requested changes should be passed on to the relevant ERO(s) for action. This means that CORE will not itself change any of its records directly in response to an elector’s notification: it will only do so in response to an update from an ERO. This means that an elector will not see an immediate change on CORE.

90. We recognise that an elector may expect CORE to instantly update their record and, indeed, that having to route the request to an ERO is more cumbersome

12 For example, the Australian Electoral Commission facility at https://oevf.aec.gov.au and the facility provided by Elections New Zealand at https://secure.elections.org.nz/app/enrol

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

than technologically it needs to be. However, we feel that to allow an ‘instant update’ facility to electors would depart from a fundamental principle set out by the Electoral Commission that the local ERO should make decisions on accepting a registration application or to change the information.

91. An elector may request a change of address when moving from an address within an area covered by a CORE scheme to an area not covered (see section on ‘Geographical coverage’ below). In this situation, we propose that the ERO for the area being moved into should be sent a notification from CORE, albeit that the process may differ slightly from that if both EROs were within the CORE scheme area.

Q18: Do you agree with the proposal that an individual elector should be able to access directly all the information held on them by a CORE system, for the purpose of confirming accuracy and/or requesting changes? If you agree, please state whether you would prefer to see the ability to confirm accuracy and request changes implemented at the same time or the ability to request changes implemented later. If you disagree, please set out your reasons.

92. The increased access to information and ability to request changes to it that the proposed individual elector access provides brings a clear need for tight security measures that verify an individual’s identity before they are given access. We recognise that if we were to require a new CORE-specific identity verification process there would be likely to be minimal take-up of an online elector access facility: a new verification process to remember - combined with the low number of occasions on which the facility would be likely to be used by any one individual - would outweigh its perceived benefit for most people.

93. We therefore propose that CORE online access is included as an additional service once individuals have registered with either the existing Government Gateway secure portal, or the developing ‘Register’ strand of the Government Connect programme (which itself will use the existing Government Gateway infrastructure).

Q19: Do you agree with the proposal that CORE should look to use the Government Gateway and/or the Government Connect ‘Register’ strand as the means by which an individual may verify themselves and gain direct online access to the information held about them on CORE? If you disagree, please set out your reasons.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Anti-fraud mechanisms

94. A major advantage of a model with a consolidated electoral registration dataset is that a range of security checks can be conducted across and between all the otherwise separate local records held. Errors, omissions, and electoral fraud can be detected and prevented. In essence, all the proposed checks search for matching records. The different types of matches are set out below.

95. The efficiency of the duplicate matching process would be very reliant on the collection of personal identifiers by EROs. Currently, in Great Britain, the only ‘personal’ information collected for every elector is their name. Nationality information is collected, but there is currently no power or requirement for EROs to store or retain that information. Similarly, date of birth information is collected only for ‘attainers’ (essentially 16 and 17 year-olds).

96. Name alone is probably insufficient for workable data matching. However, if the proposal to collect, for example, a date of birth for every elector, means greatly increasing the chance of finding genuine matching records. As personal identifiers and cross-matching security checks are already in use in Northern Ireland, useful reference can be made to how the system works there.

Different types of duplicate checks

• Elector name and/or other personal identifier.

97. CORE can check across all its elector records – including anonymous elector information – to identify where any personal identifiers match. Where matches are detected, this indicates that the same person appears to be registered at more than one address. It can be perfectly legal for an individual to be on two electoral registers. For example, university students are often entitled to be on two registers at once if their term-time and home addresses are different.

98. However, a matching entry could be an omission (e.g. the person has moved but not had their old address deleted). Such information may even reveal an attempted fraud (e.g. registering at different addresses in order to obtain multiple votes). As either possibility requires amendment to the register, local ERO(s) would need to undertake further investigation.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Q20: Do you agree with the proposal that CORE should cross-match information it holds, to identify apparent multiple instances of the same individual elector? If you disagree, please set out your reasons.

• Proxy name and/or other personal identifier.

99. Using the same type of personal information, but looking this time across proxy voter lists, CORE can detect where the same person has been appointed as a proxy more than once. It is legal for an eligible individual to vote as proxy for two different electors, but only close family members may do so for more than two electors. If it held proxy voter data, CORE could identify where the same individual appears to have been appointed as proxy by more than two electors, and thus identify where there was the greatest risk of this proxy voting restriction being breached.

Q21: Do you agree with the proposal that CORE should cross-match information it holds, to identify apparent instances of the same individual acting as proxy for more than two electors? If you disagree, please set out your reasons.

• Mailing addresses.

100. By checking across the postal voters lists and postal proxies list, CORE can detect multiple instances of the same mailing address for postal voting papers. A threshold should be set at or beyond which the system would flag up an address for investigation into possible postal voting fraud. We would welcome views on what the threshold figure should be. Local EROs will be able to quickly identify large multi-occupancy addresses such as student halls of residence or nursing homes. .

Q22: Do you agree with the proposal that CORE should cross-match information it holds, to identify apparent instances of the same address being used as the mailing address for the postal votes of multiple electors? If you agree, what do you think might be an appropriate threshold figure at which an alert should be triggered? If you disagree, please set out your reasons.

• Ballots issued/cast

101. By looking across marked registers, CORE can detect where an individual has registered to vote and received and possibly returned more ballot papers

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

than they were entitled to. As noted above, a person can be validly registered in two different ERO areas, but in a UK Parliamentary election may only vote in one of those locations. The consolidated CORE dataset could flag up for investigation and appropriate action instances of the individual being issued with – and returning – UK Parliamentary voting papers in both areas.

Q23: Do you agree with the proposal that CORE should cross-match information it holds, to identify apparently inappropriate instances of multiple voting by the same individual? If you disagree, please set out your reasons.

Action to be taken where multiple entries are discovered

• Action to be taken by CORE

102. Where matching data is identified by CORE, we propose that the keeper should send an appropriate notification to the relevant ERO(s). As noted above, a fundamental principle of the proposed CORE model is that it is not the actual register, but rather a reflection of the local register under the control of the local ERO.

103. It would therefore not be appropriate for CORE to amend the records it holds until authorised to by the relevant ERO. Any investigation into apparent duplicate information – and decisions on what action to take in light of the findings (for example to refer the matter to the police) – we believe should be the responsibility of the ERO(s) concerned.

Q24: Do you agree with the proposal that CORE should refer apparently anomalous information it receives back to the originating EROs? If you disagree, please set out your reasons.

• Action to be taken by EROs

104. Assuming the apparent anomalies are referred back to originating EROs, the question arises as to what an ERO should be required to do in response. We do not propose to specify the investigative action that must always be undertaken by an ERO. It will be relatively clear where investigation is unlikely to reveal an inappropriate elector record. For instance, where an individual appears registered in two locations and one is clearly a student hall of residence, it would be quite easy for an ERO to verify that both records are valid. 41

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

105. However, we do propose that an ERO should be required to actively respond to a notification from CORE on what to do with the records. We would suggest such a response could be either to confirm that a record is valid (and therefore should not appear in the next set of notifications from CORE), or that the matching record should be ‘held over’ and appear again in the next scheduled notification if not yet investigated. Local EROs would correct records when sending their regular information updates to CORE (see above section on ‘Compiling the record’).

106. We recognise that the requirement to respond to such notifications is likely to bring additional work to EROs and their staff once CORE is established and annually around the canvass. However, we expect the volume of work to then diminish and stabilise.

Q25: Do you agree with the proposal that an ERO should be required to actively respond to formal notifications from CORE of apparent anomalies? If you disagree, please set out your reasons. Whether you agree or disagree, we would welcome any informed assessment of what the initial and longer term resource impact of any such requirement would be on EROs.

107. A requirement to actively respond to notifications from CORE would ensure that appropriate local action was being taken. If an ERO failed to respond to notifications within the specified time period, or was consistently returning ‘hold over’ requests, this would rapidly become clear.

108. We would welcome views on how long an ERO should have to respond to notifications from CORE and how frequently they should be sent. As a large amount of notifications are likely to be generated when CORE is first established in an area (and a smaller amount around each annual canvass time), we might require notifications to be sent and responses to be made on a monthly basis. However once the system has bedded down in an area, it may be more appropriate to change the frequency to weekly.

Q26: If a requirement to respond is established, we welcome views on how frequently such notifications should be sent, how long an ERO should have to respond, and what penalty (if any) should attach to any failure to respond.

• Duplicate searches by EROs

109. So far, this section has dealt with notifications sent by CORE to EROs. However, we would also propose that an ERO can request a duplicate search

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

when considering a rolling registration or absent voting application from an elector. This would be useful in ensuring that where an elector is moving house, their previous address can be removed even if they failed to complete the ‘previous address’ section of the rolling registration form.

110. For data protection reasons, we believe this process should operate on a ‘search’ rather than ‘browse’ basis. Rather than view another ERO’s register in its entirety the ERO would first submit a specific set of elector details.CORE would respond with any information already held on that elector, or supply a nil return. We believe this process would help ensure the integrity of EROs’ records, an underlying objective of the rolling registration process.

Q27: Do you agree with the proposal that an ERO should be able to run a check online with CORE to see whether a particular individual already has a record held on that dataset? If you disagree, please set out your reasons.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Other information generated by CORE

The provision of statistical information.

111. We would also propose that the CORE generate statistical information. This could be provided to Government, academics, and the Electoral Commission for monitoring or research purposes. Clearly such data would need to be sufficiently de-personalised, so as to ensure that it did not infringe Data Protection requirements. However, in principle, we see no reason why this particular attribute of an up-to-date consolidated dataset should not be utilised.

Q28: Do you agree with the proposal that CORE should be able to provide statistical data? If you disagree, please set out your reasons.

Statutory reporting obligations.

112. CORE could also handle information required under the statutory mutual arrangements between EU member states. In order to ensure that individuals do not vote in both their state of residence and their state of origin under reciprocal obligations, each member state is required to notify the other EU member states of its residents who are those other states’ citizens and registered to vote for European Parliament elections.

113. It has been suggested recently that such arrangements may be disproportionate and no longer necessary. However, currently in the UK, the responsibility for reporting this information to other member states is left to individual EROs. Reciprocal information received from other member states is distributed to relevant local EROs via a government department. In future, CORE could handle the notification to other EU member states, providing a single consolidated notice for the whole area covered by the CORE scheme. It could also take on the task of dealing with information received from other member states and pass it on to local EROs.

Q29: Do you agree with the proposal that CORE should, once established, discharge the statutory mutual reporting responsibilities concerning EU citizens voting in European Parliament elections? If you disagree, please set out your reasons.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Establishing the first CORE scheme

Geographical coverage

• A staged approach?

114. We originally envisaged a national CORE scheme becoming operational across the UK at the same time (although possibly with a slight delay for Northern Ireland because of the different requirements for individual registration there). However, there is an argument to be made for a more staged approach to implementation.

115. CORE arrangements are technically complex and will bring a radical change to some aspects of electoral registration data-handling. It may therefore be reasonable to introduce the new arrangements in a smaller geographical area first and so limit the impact of any initial problems and resolve any problems before national roll-out.

116. We are currently open-minded on whether or not to take a phased approach to implementation of CORE national access arrangements. We therefore propose to frame the primary legislation to allow for – but not require – it.

• What might a staged approach look like?

117. Practically speaking, there would be no advantage to a first scheme applying only to an area covered by a single ERO: a single database is already used for such an area and the very purpose of CORE is to co-ordinate records otherwise held separately. Although there would be some basic testing value in having the first scheme cover just a few EROs’ areas, there is likely to be very little in the way of movement of electors and duplication of entries.

118. We therefore believe that a more appropriately sized first scheme would be – say – a whole region of England, or the whole of Wales, or at least a sizeable English county or metropolitan area. If this proved successful, the scheme might next be expanded to cover the whole of England and Wales, as the processes for electoral registration are identical. Scotland and Northern Ireland could be added as a final phase, once the differences in registration policy and processes can be worked through.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

119. We would particularly welcome suggestions on which areas might be best suited as the first scheme location under a staged approach. We would be keen to know if you foresee problems in operating a CORE scheme in only part of an area that happened to be holding an election.

120. Expressions of interest to become an ‘early adopter’ are actively encouraged, as we would wish to operate on a voluntary basis as far as possible.

Q30: Do you think CORE Phase 2 should be established in geographical stages, or to cover the whole UK at once? If in stages, please state what level of geographical coverage you believe would be appropriate for the first CORE scheme. Suggestions as to which particular area(s) should be part of the first scheme and expressions of interest would be particularly welcome.

The CORE ‘keeper’

• Who should run the CORE central infrastructure?

121. The Government’s view is that this should largely depend on the size of area covered by the scheme. If the area covered was only one region (or smaller), then the best ‘keeper’ may be an experienced local ERO (a current practitioner with front-line experience and familiarity with the geographical area covered). As the role would consist largely of oversight, we would not regard it as particularly burdensome to EROs: the substantive operational work would be handled by dedicated staff on the keeper’s behalf. Your views on EROs taking on the role of keeper would be very welcome.

122. Our vision is of a CORE scheme covering the whole of the UK. Once it covers multiple regions and in particular when it crosses national boundaries it would seem more appropriate for responsibility for CORE to lie with a UK-wide body. If starting with a local ERO as CORE keeper it is vital that keeper responsibilities and infrastructure transfer seamlessly from one CORE keeper to another as roll-out progresses.

123. Our view is that the role of keeper of a larger-scale CORE scheme should fall to the Electoral Commission, as they have themselves recommended. However, we would expect internal safeguards to clearly define the roles, responsibilities and access to information of staff in the Commission, given that

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

they would be both data ‘keeper’ and consumer (in fulfilling their statutory role in checking donations to political parties).

Q31: Do you think that the Electoral Commission should be appointed as the keeper of a CORE scheme? If you do, would your answer differ if a first scheme was on a relatively small scale? If you would prefer to see some other body appointed as CORE keeper, please say who and set out your reasons.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Future possibilities

124. Not under the terms of the current CORE project, but in the longer term, further enhancements to the comprehensiveness and integrity of electoral registers are possible. The possibilities and associated drawbacks are discussed here in order to gather initial views from stakeholders.

125. Many of the possible additional functions outlined below involve data sharing – either one-way or two-way – with other non-electoral datasets. There are therefore clearly also serious data protection issues to be worked through substantively

126. Accurate and effective data-matching across different datasets requires some form of personal identifier held in common. An immediate obstacle is that no such common personal identifier currently exists within electoral registers, other than the name itself. This is of very little use unless combined with other – more unique – personal identifiers.

Status updates

127. One simple way in which links with other databases would bring value for the electoral register is the familiar concept of the ‘one stop shop’ for the citizen. Where the citizen updated their personal or address details with another public sector database (e.g. the Inland Revenue or the Benefits Agency), that change of relevant elector status could be automatically notified to CORE for onward transmission to the relevant ERO.

128. Many citizens already believe this happens to a greater or lesser extent, and when they find it does not, often feel it should, so as to save them the inconvenience of updating their details with multiple organisations separately. A wide-ranging network of links between databases is well beyond the scope of the CORE project or DCA itself to achieve.

129. However, a particularly useful dataset with which CORE might establish at least a bilateral relationship of this sort in future may be the proposed England

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

and Wales-wide register of births, deaths and marriages13. Such a link could yield very useful information on changes of name (following a marriage) and deaths of electors. EROs and local registrars already share information in some areas but a link between national datasets would make this arrangement more consistent and comprehensive.

Verification checks

130. Electoral registration remains heavily dependent on trust. The checks currently conducted on individuals applying to register are relatively limited, if they exist at all. In the absence of automatic duplicate checking such as that proposed for CORE above, investigation of applications depends on whether somebody actively objects to an individual’s inclusion in the register.

131. Through the Electoral Administration Bill the government is seeking to tighten up certain aspects of the process. For example, the Bill was drafted to allow objections to registration to be made at any time (rather than just during the five-day period that currently applies). , to permit EROs to instigate an objection themselves and to remove inappropriate entries from the register as well as decline an application to go onto a register. The Bill acknowledges that personal identifiers can enhance the integrity of the registration process.

132. With a CORE consolidated dataset, it should be possible to check elector records against a dataset requiring much higher levels of verification. The other dataset might also make notification of changes to personal details or addresses a requirement and discrepancies could be referred back to an ERO for investigation.

133. Verification services were considered in detail in the ID card register proposals14. The anticipated high level of security checking and intended requirement for citizens to notify changes may make the ID card register dataset a particularly useful comparator.

13 http://www.gro.gov.uk/Images/01chapters1-11_tcm69-3577.pdf

14 http://www.homeoffice.gov.uk/passports-and-immigration/id-cards/how-the-data-will-be-used

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Finding gaps in the register

134. The Electoral Administration Bill seeks to introduce a new duty on EROs to maintain their registers with the aim of getting onto the register as many eligible voters as possible. Looking at registers held for other purposes is key and EROs already have the power under existing regulations to look at other records held by their own local authority, or those held by local registrars, and many already do so.

135. However, linking CORE to other datasets at a national level may identify even more gaps in the electoral register. The elector records could be matched against a different dataset in the same geographical area. Entries on that other dataset not appearing on the electoral register could be referred back to the ERO in order to maximise the number of potential electors. For instance, where another dataset revealed residents in a particular address but the CORE record has nobody listed, this could be fed back to the relevant ERO. The aim of this would be to help the ERO better target their work to get all eligible people onto the voters’ register.

136. Other datasets will vary in their value from an electoral perspective. A comparative dataset that compels people to record and update their details will be more useful than one that doesn’t, as will one that imposes high verification requirements on individuals wishing to register. Where both higher levels of verification and compulsion apply, there may indeed be an argument for a local ERO to automatically register the individual.

137. The nature of the supplementary information held by comparable databases is likely to determine whether there is a real benefit in sharing information. For example, the voting franchise is defined by nationality so any dataset that does not record nationality might trigger a significant amount of wasted work: individuals identified by a data-matching exercise may subsequently turn out to be citizens of countries other than the UK, Ireland, or member states of the EU or Commonwealth. In such circumstances the amount of pointless effort may outweigh the perceived benefit.

138. We would therefore welcome views not only on whether there should be linkages between CORE and other databases in future, but also what the most appropriate databases for CORE to link to might be.

Q32: Do you think any of these potential future linkages of an electoral registration dataset with other datasets should be explored more actively in

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

future? Please set out the reasons why you think they should or should not be further explored, and detail which datasets may be more suitable for CORE to link to if there were to be such linkage in future.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Appendix A:

Six models of national access arrangements (extracted from Accenture report)15

139. Identifying potential solutions entails pinpointing “differentiating factors.” Six independent value drivers were first identified that if delivered against would drive success for CORE Phase II. These drivers led thinking on the development of differentiating factors, by considering their practical implications on system design.

Definition of value drivers

140. Analysis of the fundamental aims of CORE Phase II led to the development of six value drivers against which CORE should deliver to be a success.

• Immediacy – degree to which voter data is up to date

• Accessibility – extent to which those who want to use the registration data can access it potentially by increasing both the number of channels and the level of automation with which users can access register data

• Comprehensiveness – proportion of eligible voters registered, potentially by CORE opening additional channels

• Integrity – degree to which voter data is internally consistent (particularly with respect to avoiding duplication)

• Flexibility – extent to which the currently disenfranchised could be provided for by opening additional registration or voting channels, such as the telephone or Internet

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

• Connectivity – ability to connect voter data to other Government departments for the purposes of verification and change identification

141. Assessment of these drivers leads to the development of differentiating factors that can be used to imply solution spaces.

First differentiating factor – Data resolution location

142. Fundamental to the integrity and immediacy of the data is the solution of integrity exceptions such as the resolution of duplicates on a national basis. Even more important for the purpose of differentiating between solutions, is the question of where the issue of integrity is resolved and the way in which Local Authority can compare and resolve their data sets. There are two approaches to this system improvement, which divide the solution space in half:

• Local data resolution, in which each Local Authority system is connected to all other Local Authority systems via a ‘network’ to share transactions and identify resolution problems

• Central data resolution, in which comparison of LA voter information against a central database identifies resolution problems

143. (It should be noted however, that the implied technical solution to central data resolution, i.e. a central database, would have a significant impact on accessibility).

Second differentiating factor – Data ownership location

144. If the data resolution is conducted centrally, it is necessary to consider a differentiating factor driven by connectivity and flexibility: data ownership location:

• Local data ownership, where “official” registers are owned (and hence supplied) by Local Authorities and the central “copy” is mainly held for system integrity purposes

• Central data ownership, where the central database is considered to be the “official” register from which data is supplied (under this model it would be necessary to devise a compensation mechanism to redirect, appropriately, revenue from the sale of the register to Local Authorities )

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Third differentiating factor – Data transaction location

145. If data is both resolved and owned centrally, the location of the data transaction becomes a key differentiating factor that is relevant to all drivers. There are two differing options for dealing with data transactions centrally rather than using the current localised systems:

• Local data transactions, where transactions are performed on current Local Authority systems which in turn replicate themselves to the central one

• Central transactions, where transactions are performed at local level on terminals connected directly to the same single, central system

Fourth differentiating factor – Resolution message direction

146. With local data resolution (as described above), there is a practical trade-off that relates to the manner in which resolution issue messages are directed between the Local Authorities:

• Undirected messages, in which all messages are sent to all Local Authorities at all times, regardless of relevance

• Directed messages, in which some intelligence in the connecting web directs messages only to the Local Authorities to which they are relevant

The solution options (models)

147. The tree diagram below (Figure 3.2.6) maps out how the application of each of the differentiating factors produces six possible solutions (or models). Each of these models is then summarised.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Summary description of models

• Model 1 – Registration Message Broker

The Registration Message Broker model has distributed Local Authority (LA) systems retaining their own data. The LA systems interact with a centralised “message broker” that coordinates transactions between them. The LA software packages are modified to provide notification to the message broker of transactions that could impact other LAs e.g. an elector moving from one LA to another. The Registration Message Broker then alerts the relevant LAs with messages that can be locally processed. No data is retained centrally under the Registration Message Broker model.

LA3 LA4

Message broker LA1 LA2

Registration Message Broker Model

It should be noted that because the messages are undirected the Registration Message Broker can generate substantial messaging traffic. As a result, the model necessitates substantial communications bandwidth at each LA and

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

substantial manual analysis to filter the messages for relevance. Duplication of validation transactions, wherein all LAs would need to be contacted, is onerous. National data access is not facilitated per se however use of the messaging technology to request data from LAs does provide some benefits to national users.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

• Model 2 – Indexed Registration Search Engine

The Indexed Registration Search Engine is designed to address the high volume of messages associated with the registration message broker. Under this model, a centralised search engine maintains a central index that enables messages about transactions to be directed to those LAs that are most likely to find them relevant. The LA software packages are modified to interact with the search engine rather than each other. The search engine first uses the messages to further build the index and then uses the index, to identify to which LAs the messages should be forwarded by cross reference between the individual details and its physical address.

This approach goes some way to address the option 1 issues relating to bandwidth and manual filtering. However, it still does not provide full national data access.

LA4 LA3

Message director LA1 LA2

Indexed Registration Search Engine Model

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

• Model 3 – Central Warehouse Distributed Transaction

The Central Warehouse Distributed Transaction model is characterised by a central data warehouse compiled from the messages sent between LAs in either of the previous two options. Relevant data are captured from these messages to update data in the warehouse before they are forwarded out to the Local Authorities.

This strategy provides a real time warehouse of national data without requiring LA systems to provide periodic batch loads of data. Once the data warehouse is established by copying in all the data currently held in LAs, it can be updated with message data only, substantially reducing network traffic, outages for loading into the data warehouse and load on the LA systems.

The Central Warehouse Distributed Transaction model should not impact performance since updates to the data warehouse are applied asynchronously. However, performing mass data synchronization transactions is not feasible under this model. There is no ongoing comparison between data in the warehouse and data held in LA systems. As such, although this is the first of the model to provide access to a national database, the database integrity would not be optimal.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

• Model 4 – Central Integrity Reporting Repository model

Unlike option 3, which only captures messages sent between LAs, the Central Integrity Reporting Repository checks for integrity across the distributed data system, including all LA systems and a central registration reporting data warehouse. By holding a subset of the voter registration data held in LAs and comparing it against transactions made in LAs, the Central Integrity Reporting Repository produces reports that identify potential validation or data integrity issues in LA databases. For example, if multiple LAs provide data that a voter is registered, the reporting repository would provide the LAs with a report of the potential duplication. The LA personnel would use the report and work together to resolve the discrepancy. For example, if a registering elector’s details share sufficient similarity to an elector already registered in another locality, the Central Integrity Reporting Repository would upload the information and either provide a warning if the elector is supposedly registering for the first time or trigger a message to the old LA that the voter is now registered in a different LA.

Crucially, although the Centralised Integrity Reporting Repository stores all data provided by the LA systems it does not enforce data integrity rules at the LA database level. As such, the data that has been provided to the Central Integrity Reporting Repository does not become official data of record for registration information. This remains in the LA databases as resolution of discrepancies is at their discretion. There does not, therefore, need be a “real-time” connection between the LA systems and central registration reporting database. This could be done using some kind of batch system, from daily dial-ups to weekly DVDs.

The Central Integrity Reporting Repository is the first solution in which national access is granted. It should be stressed, however, that national access would not be to a database with complete integrity as not all discrepancies will necessarily have been resolved.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

• Model 5 – Central Registry

The Central Registry is characterised by the creation of a centralised registration database that becomes the single source of information about the official registry status of an individual. In this solution, the Central Registry enforces integrity and only accepts data from LA systems once discrepancies are resolved. Although the LA systems continue to perform transactions, they are constantly connected to the Central Registry so that discrepancies can immediately be flagged for resolution. In the previous example of a registering elector that is similar to an elector already registered in another LA, the Central Registry immediately notifies the staff in relevant LAs and does not permit the transaction until the discrepancy is resolved.

The Co-ordinated Online Record of Electors (CORE) Consultation Paper

• Model 6 – Central System

The Central System is characterised by a single central Government registration system that maintains registration data stored in a central database. Under the central system model all LAs connect to the Central System and access their registration data via the central database - in effect the LAs just house ‘dumb terminals’ connected to the Central database. The Central System provides consistency in data validation, data integrity, and registration processes at the national level.

67 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Appendix B:

Diagrammatic representation of prospective relationship between CORE and other systems

Electoral Political ONS and other Credit public sector Commission Parties databases Reference Agencies

1 1 2 1

Co-ordinated Online Record of Electors Individual 4 Householder 3 elector (CORE)

5 5 5 5

Electoral Electoral Electoral Registration Registration Electoral Registration Officer Officer 6 Registration Officer 6 Officer 6 7 7 7

8 Returning Returning Returning Officer Officer Officer

10 9 9 Printer e-Voting software

68 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Key to labels on Appendix B

1 = Provision of copies of the register from CORE keeper to authorised users of such data. May also include direct online (tightly restricted) access into the CORE dataset for authorised users.

2 = As for 1 above, but longer term may also include: a) receipt by CORE keeper of updates on the status of individual electors from other databases; and b) data matching between CORE and other datasets to improve integrity and/or comprehensiveness of electoral register information.

3 = Online/telephone access to the CORE dataset by an individual elector, to allow them to check the accuracy of all information held on that individual by the CORE system and confirm accuracy or request changes, as appropriate. Access via provision of pre-registered security details (e.g. via Government Connect/Gateway portal).

4 = Online/telephone access to the CORE dataset by a householder, to allow confirmation of ‘no change’ to currently registered details. Access via provision of log-in details printed on pre-completed canvass form.

5 = Provision of electoral register information (and related information such as absent voter lists) and updates, from ERO to CORE keeper. Notification from CORE keeper to ERO of potential anomalies in data provided, and ERO responses to those notifications. Forwarding of information from electors under 3 and 4 above to ERO by CORE keeper. Longer term, potential for forwarding of any information derived from other public datasets under 2 above from CORE keeper to ERO.

6 = Rolling registration process passing details of individual electors between relevant EROs.

7 = Provision of the register to be used to conduct an election, from the ERO to the returning officer for the election. NB: the CORE project aims to ensure standardised structure of data output from ERO software, but its scope does not include responsibility to see that any recipient system – whether directly involved in the running of an election or otherwise – is configured to properly import the data.

8 = As for 7 above, but illustrates that the same ERO may need to be supplying data to two different returning officers (even in the same election).

9 = Provision of electorate data from a returning officer to a printer for the printing of polling cards, postal voting packs, etc.

69 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

10 = Provision of electorate data from a returning officer to e-voting software to run any e-voting aspect of an election (e.g. in an election pilot scheme).

70 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Questionnaire

We would welcome responses to the following questions set out in this consultation paper.

Q2: Do you agree that all information that currently appears on the full register of electors should be held in the CORE? If you disagree, please state which information from the full register that you believe should NOT be held in CORE, and set out your reasons.

Q3: If responsibility for retention of the marked register is in future to be the responsibility of EROs, what is your estimate of the cost of getting such information imported to a consolidated CORE system, and do you think it would be good value for money for the benefit to the integrity of the election that would result?

Q4: Do you agree that information set out in the statutory absent voter lists maintained by EROs should be supplied to a consolidated CORE system? If you disagree, please set out your reasons.

Q7: Do you agree that the Government should make the use of EML- compliant software mandatory on EROs by the end of 2006 to enable complete UK coverage of CORE Phase 1 electoral registration software standardisation? If you disagree, please set out your reasons.

Q8: Do you agree that the Government should actively pursue the possibility of using the Government Connect network for CORE data transactions,

71 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

whilst also – for the time being - exploring the viability of alternative networking approaches? If you disagree, please set out your reasons and what approach to establishing a suitable network you would prefer.

Q9: Do you agree that EROs should send updates to a CORE central system on a daily basis? If you disagree, please set out your reasons and what frequency of updating you would prefer to see instead.

Q10: Do you agree that data sent by EROs should go straight into the CORE record, with subsequent integrity checking and reporting of possible anomalies? If you disagree, please set out your reasons.

Q12: Do you think that, in areas where a CORE scheme is operational, smaller-scale users of electoral registration data should: a) only be able to obtain copies of the information from CORE; b) have the option to obtain the copies either from their local ERO or CORE; or c) only be able to obtain the copies from the local ERO? Please set out your reasons.

Q15: Do you agree with the proposal that authorised bodies should be granted direct electronic access to the CORE central dataset to browse and/or initiate an electronic search for an individual record. If yes, we would welcome your views on how a charging structure might work for those who are normally expected to pay for copies of the register. If you disagree, please set out your reasons.

Q16: If direct access to a CORE system for the purpose of confirming identity were to be established, we welcome views on who exactly should be given the ability to use that facility.

72 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

73 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Q24: Do you agree with the proposal that CORE should refer apparently anomalous information it receives back to the originating EROs? If you disagree, please set out your reasons.

Q25: Do you agree with the proposal that an ERO should be required to actively respond to formal notifications from CORE of apparent anomalies it has detected? If you disagree, please set out your reasons. Whether you agree or disagree, we would welcome any informed assessment of what the initial and longer term resource impact of any such requirement would be on EROs.

Q28: Do you agree with the proposal that CORE should be able to provide statistical data? If you disagree, please set out your reasons.

Q32: Do you think any of these potential future linkages of an electoral registration dataset with other datasets should be explored more actively in future? Please set out the reasons why you think they should or should not

74 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

be further explored, and detail which datasets may be more suitable for CORE to link to if there were to be such linkage in future.

Thank you for participating in this consultation exercise

75 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

About you

Please use this section to tell us about yourself

Full name Job title or capacity in which you are responding to this consultation exercise (eg. member of the public etc.) Date Company name/organisation (if applicable): Address

Postcode If you would like us to acknowledge receipt of your response, please tick this (please tick box) box Address to which the acknowledgement should be sent, if different from above

If you are a representative of a group, please tell us the name of the group and give a summary of the people or organisations that you represent.

76 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

How to respond

Please send your response by 07/03/2006 to:

Xavia Morbey Department for Constitutional Affairs Electoral Policy Division 6th Floor Selborne House 54-60 Victoria Street London SW1E 6QW

Tel: 020 7210 8302 Fax: 020 7210 2659 Email: [email protected]

Extra copies

Further paper copies of this consultation can be obtained from this address and it is also available on-line at http://www.dca.gov.uk/index.htm

Publication of response

A paper summarising the responses to this consultation will be published by the end of June 2006. The response paper will be available on-line at http://www.dca.gov.uk/index.htm

Representative groups

Representative groups are asked to give a summary of the people and organisations they represent when they respond.

Confidentiality

Information provided in response to this consultation, including personal information, may be published or disclosed in accordance with the access to

77 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

information regimes (these are primarily the Freedom of Information Act 2000 (FOIA), the Data Protection Act 1998 (DPA) and the Environmental Information Regulations 2004).

If you want the information that you provide to be treated as confidential, please be aware that, under the FOIA, there is a statutory Code of Practice with which public authorities must comply and which deals, amongst other things, with obligations of confidence. In view of this it would be helpful if you could explain to us why you regard the information you have provided as confidential. If we receive a request for disclosure of the information we will take full account of your explanation, but we cannot give an assurance that confidentiality can be maintained in all circumstances. An automatic confidentiality disclaimer generated by your IT system will not, of itself, be regarded as binding on the Department.

The Department will process your personal data in accordance with the DPA and in the majority of circumstances, this will mean that your personal data will not be disclosed to third parties.

78 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

The Consultation Criteria

The six consultation criteria are as follows:

1. Consult widely throughout the process, allowing a minimum of 12 weeks for written consultation at least once during the development of the policy.

2. Be clear about what your proposals are, who may be affected, what questions are being asked and the time scale for responses.

3. Ensure that your consultation is clear, concise and widely accessible.

4. Give feedback regarding the responses received and how the consultation process influenced the policy.

5. Monitor your department’s effectiveness at consultation, including through the use of a designated consultation co-ordinator.

6. Ensure your consultation follows better regulation best practice, including carrying out a Regulatory Impact Assessment if appropriate.

These criteria must be reproduced within all consultation documents.

79 The Co-ordinated Online Record of Electors (CORE) Consultation Paper

Consultation Co-ordinator contact details

If you have any complaints or comments about the consultation process rather than about the topic covered by this paper, you should contact the Department for Constitutional Affairs Consultation Co-ordinator, Laurence Fiddler, on 020 7210 2622, or email him at [email protected]

Alternatively, you may wish to write to the address below:

Laurence Fiddler Consultation Co-ordinator Department for Constitutional Affairs 5th Floor Selborne House 54-60 Victoria Street London SW1E 6QW

If your complaints or comments refer to the topic covered by this paper rather than the consultation process, please direct them to the contact given under the How to respond section of this paper at page 77.