Electoral Registration Data Schemes (No. 2) Order 2012

Lord Rennard Excerpts

Tuesday 27th November 2012 Grand Committee

[Lord Wallace of Saltaire]

My Lords, The order will provide the legal basis for a further electoral registration data-matching trial, by enabling the sharing of specified data between several data-holding public authorities and some 22 local authority electoral registration officers. The work that we plan to do under this order will form a significant part of our planning for the implementation of individual electoral registration.

The Committee will be aware that this is the third draft order of its kind since the summer of last year. It may be helpful if I were briefly to recap the story so far, before I go into detail about what the present order will do.

The first order—the Electoral Registration Data Schemes Order 2011—allowed us to carry out a set of data-matching schemes and evaluate the results. In those schemes, we were trying to find out whether matching their registers against public authority databases would help electoral registration officers to find potential electors who were missing from the register, so that they could contact those people and invite them to register. We were also trying to find out whether data matching would help registration officers to find entries on their registers that might be inaccurate or fraudulent, so that they could investigate them and then, if necessary, take steps to remove them.

We learnt a lot from those first schemes about the challenges of data matching and about the techniques and the technology that we would need to put in place if we were to do data matching more effectively and on a larger scale. When the schemes were over, the evaluations told us that more piloting work would be needed if we were to ascertain the potential of data matching, and data mining, for finding potential electors who are missing from the register.

What last year’s schemes did reveal, however—rather unexpectedly, it is fair to say—was that data matching might give us a way of confirming the majority of existing electors on the register in the transition to individual registration. If that turned out to be correct, most of the electorate would not have to register individually as soon as individual registration is introduced. That would be more convenient for electors; and for electoral registration officers. There would be significant savings in time and money which would enable EROs to concentrate on those whose details could not be matched and those who were missing from the register altogether.

We needed to test our understanding, however, and we needed to do it quickly so that, if this was shown to work, the necessary systems could be put in place in time for the transition. The second order—the Electoral Registration Data Schemes Order 2012—approved by the House in the summer, was sought mainly to enable that testing, and the schemes for confirming existing electors are now in progress. That order also allows us to carry out further testing of data matching for finding missing potential electors and inaccurate or fraudulent entries; but only in the areas specified in the order, and only using data held by the Department for Work and Pensions. But I told the Committee in the summer that if we decided to extend the schemes to include further areas or data sets, a separate order would be laid before your Lordships at a later date. We are now ready to do that further testing, hence the order now before the Committee.

This latest order will allow EROs in the areas listed to compare their registers against specified public authority data sets. The public authorities which have agreed to make their data available for these schemes are the Department for Work and Pensions, for areas that were not included in the previous order; the Department for Education or, for schemes in Wales, the Welsh Government; the Student Loans Company; and Royal Mail Group. The schemes will target three particular groups where there are high levels of under-registration: people who have recently moved home; young people of 16 to 18 years of age who are just going on to the register; and students. They will complement a programme of work that the Cabinet Office has in hand to maximise electoral registration among groups identified as currently under-registered and at risk of falling off the register during the transition to individual registration.

The main purpose of the schemes will be to see how far data-matching helps EROs to improve the accuracy and completeness of the register by finding people who are missing from the register and finding entries on the register that should not be there. The schemes will also help us to design, develop and test the technology that we will need if data-matching is to play a significant part in future arrangements for electoral registration.

In addition, the order will enable EROs in four lower-tier authorities in two-tier local government areas to match their registers against education data held by their county council, to see whether it helps them to find 16 to 18 year-olds who are not yet registered. Registration officers in unitary authorities already have access to such data because it is held by the same authority that appointed them, but their counterparts in two-tier areas have no right to access the same kind of data if it is held by a different council. The results of these schemes will help us to decide whether it would be worthwhile to legislate to correct this anomaly.

The order will also enable us to augment the work that is already being done on confirming existing electors, by allowing us to carry out a statistical analysis to find out how far other public authority data sets might add to the match rate obtained from DWP data. The Higher Education Funding Council for England has agreed to make its data available for this purpose alone.

As in previous instruments of this kind, the draft order requires that before any data can be transferred, a written agreement must be in place between the ERO and the data-holding public authority setting out the requirements for the processing, transfer, storage and destruction of the data. It sets 17 July 2013 as the date by which each of the schemes must have been evaluated by the Electoral Commission. I also assure the Committee that after the pilots have ended and the evaluation is complete, the data created and held for the purposes of the pilot schemes will be securely destroyed.

The Information Commissioner’s office has been consulted on this draft order. The office has welcomed the fact that the current phase of pilot schemes has identified a much narrower range of data, and that the schemes will inform the extent to which personal data to be collected from electors can be minimised. I hope that the Committee will recognise the value of this further work for improving the accuracy and the completeness of our electoral registers.

I would like to make some additional points which I hope will help the Committee. All of this feeds into the wider context of digital transformation and the development of what in the trade is called “identity assurance”. This morning, I had a useful briefing from the government digital service on exactly this matter. Further down the road, there are delicate issues about the balance between the use of public and private databases, to which we will want to return in that wider context. I reiterate that the current electoral register has deteriorated quite badly over the last 25 years —especially in its coverage of vulnerable groups. We are very conscious of that and are therefore strongly committed to this move toward individual electoral registration and to using this transformation to maximise the accuracy and completeness of the electoral register. I hope that the Committee will accordingly approve this order.

[Lord Rennard]

These orders are clearly welcome to me, as they show potential ways forward for increasing voter registration and improving the accuracy of the electoral register. It seems to me, however, that further pilots are clearly needed, because earlier pilots were certainly not considered to have been a complete success. The Electoral Commission raised serious concerns about the reliability of the earlier pilots because they had,

“an absence of a clear, common, methodological framework”.

This, it said,

“had a significant impact on our ability to draw clear conclusions about the effectiveness of data matching as a tool for maintaining the accuracy and completeness of the electoral registers”.

The Electoral Commission has raised a number of concerns about this next set of pilots, of which, I am sure, the Minister will be aware. In particular, will he tell us how closely the IT systems to be used in these pilots will match the IT systems being developed for eventual use in implementing individual electoral registration? It is clear that they are not the same systems, as the eventual IT systems to be used are not yet ready. Does he therefore accept that there is a significant element of risk in making an assessment of these pilots and drawing conclusions about the effectiveness of the IT systems that will eventually be used?

The commission raised a number of concerns about the methodological framework for the pilots. I am sure that the Minister will assure us that the Cabinet Office will do its best to address them. He has told us that the commission will evaluate these pilots by 17 July 2013. It seems to me that the crucial issues of the completeness and accuracy of the electoral registers will depend on the relative success or failure of approaches being taken in these pilots and other measures which are yet to be announced. It certainly will not be before these pilots can be evaluated that we will know whether a register based exclusively on individual electoral registration will be fit for purpose. That is why the existing Bill must provide for Parliament to decide whether the process has been sufficiently successful for our elections and for future boundary reviews to rely exclusively on it; just as Parliament will also have to approve any decision to abandon the annual canvass.

When the Minister responds, I hope that he will provide some clarity to the Committee about when the Bill will come back and we can debate further the issue of when it may be considered safe to rely exclusively on an electoral register based on IER. In the mean time, we have to hope that the transition will be as successful as possible, as quickly as possible, in terms of the stated aims of improving the completeness, as well as the accuracy, of the electoral register. It seems to me that these aims are best served by testing as many potentially relevant databases as possible. Use of the DWP database will help to ensure that, for example, people who are retired will be registered. The DWP is clearly happy for its database to be used in that way.

However, I understand that the Department for Transport has not given permission for its database at the DVLA to be used in a similar fashion. Both databases are national, government databases and both, of course, will have significant levels of inaccuracy. Surely, it would be better to use them both rather than just one of them. Perhaps the Minister will explain if the DVLA database will be used in due course. It would be very disappointing and quite unacceptable if the Government, having been asked repeatedly to use the DVLA records, were to argue in the future that the fact that there had been no pilots with the DVLA data was the reason for not using the DVLA database for the final process of transition to IER. The DVLA holds data on millions of adults, which is reasonably up to date, because it is a legal requirement to notify the agency if you move.

I very much welcome the addition to the list of databases secondary schools and academies, the Higher Education Funding Council for England, the Student Loans Company and the Royal Mail Group. The presence of educational institutions makes particular sense when it comes to adding so-called “attainers”; that is, young people who are coming up to voting age. I hope that the presence of those institutions in this list is an early sign that the Government will accept that the use of secondary schools’ pupil information must be integral to the IER regime, as it is in Northern Ireland.

It seems to me that in this respect we have at least had a four-year pilot in Northern Ireland. My understanding is that it has been very successful in engaging with 16 and 17 year-olds to add them to the register. We learnt today from the report of the Electoral Commission on registration in Northern Ireland that it was probably unwise to abandon the annual canvass there.

--- Later in debate ---

[Lord Wallace of Saltaire]

I am very happy to take that back. I will report back to my colleagues on the strongly held sentiments. Perhaps I may take the questions about tenancy and deposit schemes and credit agencies at the same time. The initial assessment by the Cabinet Office of the suggestion from my noble friend Lady Berridge that tenancy deposit schemes might be used was that it was not sufficiently obvious that the processes of these databases could be adapted to support IER. However, that does not exclude renewed consideration.

Of course, the question of credit agencies takes us over the boundary between public and private. Credit agencies are part of the private sector. The issue is part of a broader discussion that we all need to have with the likes of the noble Lord, Lord Maxton, and others, about the extent to which, as we move into a new world of data transmission and availability, private and public databases can be used for identity assurance. That was the basis for the briefing I received this morning from the government digital service. It would be helpful to organise a meeting for Peers as a whole on the work that it is doing—for longer-term and wider purposes than this Bill alone—on these issues. Private databases are increasingly useful, but their use raises questions about civil liberties and public and private interests with which we need to be concerned.

[Lord Rennard]

The Minister suggested that there was a great gap between private sector credit reference agency databases and public sector databases. Would he not accept that private sector databases used by credit reference agencies are already used extensively by public local authorities? Many local authorities use data held by credit reference agencies to determine whether there may be more than one person living in a household, in particular when someone is claiming a single person’s council tax discount. Credit reference agencies may have information suggesting that more people are present in the house, and revealing who they are. Local authorities, which are public sector organisations, are already using the data from private sector credit reference agencies. Would it not be logical for electoral registration officers to do what their colleagues in finance departments are doing to identify the existence of people who are there but who are not on the electoral register, and invite them to be on the electoral register? I am not aware of any objections from civil liberties groups to any of these existing practices.

[Lord Wallace of Saltaire]

I thank my noble friend for that strongly worded intervention. I take that on board as one of the issues that we are edging towards. The civil liberties lobby may not have caught up yet with the point that he is making, but I expect that it will do so soon. There are some very broad issues here that we have to be concerned about. I point out, as he has done, that one of the principles of our system of electoral registration is that it is in the hands of local authorities. We do not have a central database, so what one local authority does with credit agencies may be rather different from other local authorities do.

On the question of why this particular collection of local authorities was chosen, the answer is that these are the ones that volunteered to take part. They seem to us to be relatively representative, but this is the nature of the system under our current legal arrangements. Happily, the selection of local authorities is sufficiently wide that we and the Electoral Commission are persuaded that they will provide us with sufficiently reliable information.