It is nothing new for a government to be talking about improving electoral registration by matching data between different sources. What is different about Nick Clegg’s latest comments on the subject, in a speech to the Hansard Society, is that the talk is now becoming much more specific, with pilot projects starting next year.
Local councils, for example, hold name and address information about people in several different databases. If they are able to make use of the data from other sources to highlight either gaps in the register, or suspicious entries that may be the result of fraud, this could improve the quality of electoral registers.
Many details are not yet sorted (for example, if a missing entry is identified, could they be added to the electoral register, perhaps with notification to the person, or would the information be used to send round a member of council staff to try to persuade the missing person to register?). In addition, there are important issues around data protection and privacy, although given the Information Commissioner’s willingness to criticise the government over similar questions in the past, it is likely that the pilots will be drawn up very carefully in this regard.
Perhaps more importantly, there is not yet much sign of a fundamental shift in the emphasis on sharing data – for asking, “What data can the public sector share with itself?” to asking, “How can we make it easy for a person to let us know what data they would like to be shared?”. In other countries, for example, someone filling out their name and address on one official form can also opt in to have the information shared with the electoral registration authorities. That brings benefits of joined-up data whilst keeping control in the hands of the people whose data is being shared.
Follow @LibDemVoice on Twitter
Like us on Facebook
Subscribe to our feed
Sign-up for our daily email digest
The "Just War" tradition
Our kids need a lorra, lorra, larffs!!
West End Ward Tuesday Surgeries #dundeewestend
The Joy of Six 1275
Find Halloween and seasonal gifts in my Etsy shop today.
Mick Taylor
Pamela Manning
Simon R
Nonconformistradical
18 Comments
Data matching to get people to register to vote? It’s hardly liberal.
So the Liberal Democrats have done another U-turn. I thought you lot didn’t approve of large centralised databases? Personally I don’t have much of an opinion either way, as long as it’s kept out of private hands and is secure, but this is something that the Lib Dems were against before they were intoxicated by a little sniff of power.
This isn’t about having big centralised databases, this is about having the various databases we already have talk to each other, so that public servants can do their jobs better & so we can use them for a wider variety of tasks. Which should actually negate the need for a big centralised database…..
G: As I pointed out in the post, this can be done in good or bad ways when it comes to personal data and privacy. Given you’ve reached the conclusion in your comment, you presumably know further details of what is being planned – in which case, would you care to share them?
Of course, if you didn’t, then leaping to the conclusion that this involves a large centralised database would look like a foolish leaping to a clichéd assumption in the absence of evidence. And I’m sure you wouldn’t be foolish 🙂
“So the Liberal Democrats have done another U-turn. I thought you lot didn’t approve of large centralised databases?”
The Lib Dems don’t approve of a large centralized database but they seem to be in favour of several small databases which are centrally linked, which, in turn provides greater access to each database.
(yeah I know what your thinking, that’s what I thought as well, go figure)
Mark Pack, let’s say I have multiple datasets on my computer. Let’s say I want to easily analyse and compare each dataset. Do I a) Make a new database containing all the data in one place and output the comparison in a format of my choice or b) write a script to pull out the relevant data from each individual database and output the comparison in a format of my choice?
The answer would be whichever option is quickest and easiest from my pov because the output will be the same.
We can argue the semantics if you like, but that would be to miss the point that once you link separate databases with a tool you have created something that gives the same answer as a single database would. Technically the former is not a single database, but at the point of use that distinction is not detectable.
“If they are able to make use of the data from other sources”
“Many details are not yet sorted”
Sounds like a great plan.
“send round a member of council staff to try to persuade the missing person to register?”
Local authorities have to make 28% cuts in spending over four years – and you think this is possible?
Pie-in-the-sky thinking. How can anyone take this seriously?
I would need to be persuaded of the wisdom of these public sector databases sharing data between them because of long-held liberal concerns about goverment use of personal data. However, having the possibility of people opting into data sharing in order to be included on the electoral register strikes me as a good idea.
Owners of small businesses may opt to stay off the register to ensure that they won’t be called up for jury service; this is usually for up to a fortnight only but if you get a longer case [some last months] then there is no escape; their enforced absence may ruin their business. Local government wants them on the register because it is used to estimate population and therefore need when allotting grants.
Are they to be pursued and persecuted to get them to sign up?
G: Your example makes my very point. That’s one way of matching data, but it’s only one of very many different ways that data matching to identify problems with the register could be done. For example, a limited anonymous subset of data could be taken from one set of data (e.g. ‘here’s a list of properties at which there is at least one adult listed’), another taken from another database (e.g. ‘here’s a list of all properties for which there is no-one on the electoral register’) and those two files compared with each other to produce a third list (‘here’s a list of houses to check for under-registration’).
That’s very different from one big centralised database, or indeed a network of linked databases.
Fiona: what’s so bad about a plan for which all the details are not yet sorted? Should people only talk about plans which appear fully formed?
@ Mark Pack
“That’s very different from one big centralised database, or indeed a network of linked databases”
I do see your point, however it is still a step in the wrong direction, yes a small step admittedly but a precedent maybe set. It may even be done for the right reasons but the path to hell is often paved with good intentions
Mark, I think the point is that once you link databases they nmeed for centralisation becomes irrelevant because the linking is an adequate substitute. What I’m trying to get at is the underlying principle here, do you object to the idea that one terminal should have access to all the knowledge?
It makes no difference whether there are linked databases or a central database, the output from the terminal is the same. There are important differences from a costs and maintenance perspective, but that’s a matter for costing and predictions of future infrastructure changes – important, but hardly matters on which principles rest. The important question is that do you think one terminal should have access to the entirety of the data? Because that is where Clegg is headed.
As I said though I have no objection here, and I think his reasoning is sound, but if the Lib Dems are still opposed to ‘centralisation’ then I think they are obliged to oppose this measure, not to pretend that linking databases is somehow different from creating a single one, because at the point of the user it’s not.
G: You talk about one terminal having access to all the data. How do you know that’s what is going to happen with this scheme? That isn’t the only way data matching is done in other circumstances and other places, so how do you know that none of those other approaches will be taken in this case? (I don’t know, which is why in the post I left my view of the pilots up in the air; it’ll depend on which of the many different ways of data matching are selected for trial.)
I doubt if any of this is feasible at a local or national level because of the cutbacks.
The quality of data in the multiple formats used for electoral registration is not good at all. It is probably the same for the council tax database and other local authority relevant data. Then matching via the post code national (Royal mail owned) database is not infallible either.
This all adds up to an awful lot of manual checking and difference resolution, probably not justified by the costs. A persuasion campaign would be cheaper.
I understand the previous gov started a national electoral database project – I wonder what happened to that?
>Local councils, for example, hold name and address information about people in several different databases.
But they don’t have info on everyone. You’d have to get info from HMRC/banks/DVLA to be sure you had a complete database. All too Big Brotherish for me.
>would the information be used to send round a member of council staff to try to persuade the missing person to register
No one home during working hours in surely millions of cases. And if some council official came to my house one evening or weekend, to ask why my son isn’t registered to vote (answer being he’s registered in his university city), I’d see that as a waste of council resources.
Cassie is right:
“But they don’t have info on everyone. You’d have to get info from HMRC/banks/DVLA to be sure you had a complete database. All too Big Brotherish for me”.
It all sounds rather like a national ID system – without the benefit of a card.
No-one will be fooled by this.
As for this:
“what’s so bad about a plan for which all the details are not yet sorted”
Well then it’s not a plan, is it! It’s just an idea, one which looks rather confused at best and naive at worst. It’s the kind of thinking you can get away with in opposition – but not in government.
You know, I could have seen exactly that quote in a Jacqui Smith statement about the National Identity Scheme. I’m no more likely to trust the Coalition with this sort of scheme than I was to trust Labour.
Doing this in the name of voter registration is idiotic. Perhaps the reason our voter registration levels are lower than they could be, despite huge public advertising campaigns, simple registration processes etc., is that people are fed up with politicians proposing exactly the type of scheme that they opposed out of Government. Let people stay off the electoral register if they’re not interested in perpetrating this charade.
Mark although the intentions are good there is a lack of oversight in these propsals. What worries me is that Clegg does not appear to have grapsed the point about personal privacy and the Database State, and fails therefore to see how it applies in this circumstance.
Clegg specifically mentioned adding National Insurance numbers to the data held, presumably to act as another unique identifier. There is a huge risk here of turning the electoral roll into a national identity register. We also have to view these proposals within the context of what the electoral commission calling for E-Voting and for requirements to provide ID and even biometric verification in order to vote.
Powers to cross-match data would overturn the fundamental data protection principle that information gathered for one purpose may not be used for another without consent. It creates a privacy minefield. Talk of anonymity amongst large networks of data-sets is now thought by many leading academics to be a unachievable due to the very powers of cross-referencing that are being proposed
NO2ID will be hoping that these concerns are listend too!