The Electoral Registration and Administration Bill currently making its way through parliament transfers responsibility for voter registration from the ‘head of household’ to the individual. This issue undoubtedly strikes many people as a purely technical, administrative exercise of interest to anoraks only. In fact, electoral registration is the lynch pin of democracy: without a fair and inclusive system for establishing electoral eligibility, the democratic mechanism cannot function effectively. For this reason, it is extremely important that the government gets this seemingly technical issue right.
The good news is that the government have rectified a major weakness in their initial proposals by dropping the previously-mooted suggestion to move from a compulsory to a voluntary system of electoral registration via an opt-out. Making registration voluntary would have considerably depressed up-take and would have resulted in a decline in what are already worryingly low levels of turnout.
But there remain two profound – and so far unremarked-upon – flaws in the government’s current plans, which could leave large numbers of voters disenfranchised. These are: the requirement that electors provide their national insurance (NI) numbers as part of the electoral registration process and the proposal to employ ‘data matching’ to compile the register.
One of the hallmarks of democracy as a system of rules is that all citizens have equal political rights, regardless of their economic means and regardless of the economic contribution they make. This is a key feature of democracy that makes it palatable to many uncomfortable with the level of economic inequality in capitalist societies. The guarantee of (formal) political equality provides the dispossessed with hope and dignity.
Any measure that erodes the cordon sanitaire that separates the political from the economic sphere represents a potential threat to democracy. The requirement to provide NI numbers as part of the electoral registration process does just that. Despite assurances from the government that NI numbers will not form part of the actual register but will only be used for the purposes of cross-checking and verifying information, people are nevertheless likely to perceive that their eligibility to vote is tied to their contribution to the workforce. Indeed, a recent Cabinet Office study entitled ‘Under-Registered Groups & Individual Electoral Registration’ found that: ‘The idea of providing a National Insurance number to register to vote was met with some apprehension and unease amongst some [research subjects].’ It went on to warn that the NI request has the potential to ‘disengage some people from the registration process’.
Under the new system, electoral authorities will also be given greater powers to use other non-electoral datasets to identify eligible electors and enter them directly onto the electoral register, a system known as ‘matching’.
Generating an electoral register from other sources of data held by government agencies is routine practice in most European democracies (which employ so-called ‘automatic electoral registration’). The UK has long had a hybrid system whereby the electoral register is compiled partly on the basis of existing information, partly from the annual canvas, and partly from information that people voluntarily supply when, for example, they move house. The use of ‘matching’ from data sources such as the Department for Work and Pensions is therefore not a dramatic departure from current practice, but it does nevertheless involve the use of data for democratic purposes that were collected as part of quite different processes. This type of approach has a chequered past. The ill-fated ‘poll tax’(community charge) introduced in 1990 under Margaret Thatcher had the (unintended?) consequence of leading an estimated 350,000 people intentionally to remove themselves from the electoral register in England and Wales for fear of being identified through electoral data as being liable to pay the tax.
If the general public become aware that data is being shunted from one branch of government to another, many may fear that data they supply to electoral registration officials could flow the other way also, potentially revealing to state agencies information they would rather those agencies not have. The recent decision by Nick Clegg to effectively axe the planned boundary revisions could well mean that in the short term less matching is required (as electoral registration officers will be able to rely more heavily on information contained in the existing register) – but under the current plans it will soon become routine.
The provision for data matching and the requirement for electors to provide NI numbers together represent a potentially significant risk to the completeness and representativeness of the electoral register. Those who fail to register for fear of how their data might be used will not be a representative selection of the population; they will undoubtedly be drawn disproportionately from the ranks of Labour and Liberal Democrat voters, as happened at the time of the poll tax.
The effect of the proposed changes to electoral registration procedures could well be highly political, and wider public debate should be initiated on these measures before they are introduced.
Undoubtedly there are many problems with the antiquated system of electoral registration that justify a switch to individual electoral registration. Done properly, such a switch could lead to a more complete and accurate register. But if the two flaws in the government’s current plans are not addressed then we will likely see rates of electoral registration fall, with potentially devastating consequences for the body politic.