I was interested by the Opposition's approach. I was more heartened at the end of the speech by Mr. Djanogly than I was at the beginning, once he had made it clear that he broadly supported measures aimed at improving the level of electoral registration. The wording of amendment No. 14 suggests that the Opposition's remit is to get people off the register rather than on to it. That was the major problem that I, along with many other Members, raised on Second Reading.
Currently, electoral registration is not fair. It is biased, and discriminates against people in inner-city areas, people in housing with multiple occupation, people who are young, people who are black and people from the Asian community. We must therefore find a better system to achieve fairer registration that more accurately reflects people's entitlement to register. I accept that people who have no entitlement should not be on the register—that goes without saying—but we must address this major problem as well.
I have a view on the moves we should be making and the long-term objective of completely revised electoral registration, and my amendments conform with that view. The annual canvass with a bit of information coming from one or two databases—such as council tax information—is a pretty inefficient way of compiling an electoral register. Over two or three months, a great deal of effort is put into going around and offering people the opportunity to provide information that is already available to electoral registration officers because the form was filled in a year earlier and nothing much has changed.
The joint report of the Constitutional Affairs Committee and the Office of the Deputy Prime Minister Committee drew on evidence that we had seen in Australia, where electoral registration information is obtained from various databases and used to follow up individuals whose circumstances have changed. By doing that, rather than conducting an annual canvass of everyone, the authorities managed to produce a register which they told us they considered to be 98 per cent. accurate.