Having written previously about the downside of the government's decision to scrap the Co-ordinated Online Register of Electors (CORE), it is worth mentioning a report published today by the Commons political and constitutional reform committee.

The bulk of the report looks at issues related to individual voter registration, but CORE gets a passing mention.

"A central electoral register, such as the one that is in place in Northern Ireland, would have made identifying duplicate entries much simpler, but in July 2011 the government decided to abandon plans for a Co-ordinated Online Register of Electors on the basis that it was not "proportionate, cost effective or consistent with the government's policy on databases and reducing the number of non-departmental public bodies".

"The Electoral Commission and Dr Stuart Wilks-Heeg, director of Democratic Audit, both told us that without a central register, identifying duplicate entries would be difficult and resource-intensive, and in some cases impossible."

It is worth noting in passing that the MPs seem to look on CORE as a tool to effectively manage the electoral register, while in contrast the government spun it as being "principally to help political parties".

Both of them, however, miss the wider point about opening up the electoral system which I covered in the previous post.

Data matching not going well

The report goes on to say that the government's alternative to a central database "relies largely on data-matching with information held by other public bodies".

But the evidence on this does not appear to bode well.

Indeed, the first question would be how much of the £11m "saving" from scrapping CORE will have to be spent on running these data-matching processes.

Quite possibly more than £11m.

Data-matching pilots are currently underway, and the government hopes to assess early in 2012 whether they should be rolled out more widely.

But the MPs said:

"Representatives of three of the local authorities involved in the data-matching pilots told us that 'for all of us... it is very, very labour intensive'. All three authorities had hired additional staff to help run data matching, boosting staffing levels in their electoral registration sections by 50-100 per cent for the duration of the pilot. If it is to be successful, additional resource will be needed not only to match data, but also to follow it up with letters, and house enquiries if appropriate."

The data from the Department for Work and Pensions (DWP) seems to be particularly unhelpful.

"Julian Bassham, electoral services manager for the London Borough of Southwark, told us that data-matching 'has been more successful for us at this stage in telling us what we do know rather than what we don't know... At the moment it does not look, from our side, like the DWP data will necessarily answer those questions'."

In Southwark a startling 25 per cent of DWP records "could not be matched to properties in the borough as known to the local authority".

Friday's report suggests the problems are partly due to data consistency and standards, with addresses be stored differently or house names having changed.

But the DWP also seems to be providing both not enough and too much data: no information on nationality and therefore entitlement to vote, but also too many unneeded records

"Without nationality information, it is impossible to know if someone is likely to be eligible to vote or not. We also heard that people are not removed from the DWP database when they die or leave the country, meaning that large numbers of records on the database are essentially inaccurate and confusing to electoral administrators."

Given such unpromising prospects, it isn't surprising that the Electoral Commission said it wishes to see "further options" for identifying unregistered electors "not only in the event that data matching is less successful than we hope, but also to deal with 'at risk' groups who are less easily picked up through the data matching approach".

Unsurprisingly, the MPs conclude:

"The evidence we have received... suggests that data matching will be of limited effectiveness, especially in identifying potential electors."