My hon. Friend is right, and one registration ombudsman, as it were, has already expressed his concerns about that matter.
On the accuracy of biometric testing, Mr. Oaten discussed the results of the Government's enrolment trial in which the best result, 96 per cent. accuracy, was obtained by iris recognition. If iris recognition were used as the sole identifier, however, it would raise the possibility of up to 2 million false records in a database with 50 million entries. The Government have got wise to that point, which is why they have decided to use three biometric indicators—the iris, facial biometrics and fingerprinting—to make the database more accurate. Even if we were really lucky and achieved a 99.9 per cent. success rate, which might sound fine, it would still mean a potential failure rate of 1 in 48,000, which means that someone could access several different records using their own biometrics by creating different identities. Again, such a system would be a honey pot for organised crime and terrorists.
In response, the Government say that that trial was designed not to test the biometric technology, but to see whether people find the process acceptable, and that they are working to develop acceptable performance. They should not have introduced the Bill before calculating acceptable performance levels and whether the scheme is possible. A 99.9 per cent success rate, which I have mentioned, is way beyond the realm of possibility. In addition, 31 per cent. of disabled participants in the trial found iris recording very or fairly difficult, while the failure rate for disabled people, black people and people over 59 was low. The Government cannot deal with those issues, because even if they improve the technology, there will still be a failure rate and therefore a way for people to obtain multiple identities.
The biometric information will be stored on a massive centralised database. The Government have discussed EU proposals on biometric passports and identity documents, but the European Commission has expressed doubts about the storage of such information on a database:
"The Biometric information and its link to other personal details must be stored in a secure manner. The issue of how and where Biometrics should be stored was of particular concern to the European Commission's Data Protection Working Party. Their publication Working Document on Biometrics addresses this by looking at whether biometric information should be kept on smart cards or similar devices retained by the individual or whether it was acceptable to store the information on a centralized database. The Working Party's clear preference is for the former as it believes centralized storage presents an increased risk of data misuse. It is also important to understand that any information linked to a biometric record is only as reputable as the security procedures associated with it. A 100 per cent. reliable biometric system is of little use if any associated information is inaccurate."
That is why biometric passports, which will include facial biometrics and fingerprints, but not a central database are being introduced in Germany, which is a crucial difference.
The Government's assurances on cost beggar belief. They say that 70 per cent. of the cost of introducing the identity scheme will be incurred anyway because of biometric passports, but those assurances are about as credible as those they gave us about the threat presented by Saddam Hussein.
In fact, all that is required is a card and the information to be stored on it. It is true that the current passport database is massive, with 44 million records, but all it contains is the information that one has to give when one applies for one's passport. The photographs are stored separately. There is no storage of biometric data. The current system is minuscule in comparison with what is proposed.
On cost, the Government have rubbished the estimates put forward by the London School of Economics. So far, the only reason that they have given for doing so is that the LSE assumes that the biometrics will need to be replaced every five years. Yet the Government have said that themselves. I have a parliamentary answer from the former Home Office Minister, my right hon. Friend Beverley Hughes, which says that
"in order to ensure that the cards, which incorporate technology features such as a microchip function, are reliable, it may be necessary to replace the cards after five years."—[Hansard, 10 December 2003; Vol. 415, c. 497W.]
In any case, the cost of collecting the data is minuscule in comparison with the cost of setting up and maintaining the database and trying to keep it as secure as possible.
Many of my colleagues have been beguiled into thinking that this is a good idea and were encouraged to send out dubious questionnaires urging people to support the party because of this proposal. Do they think that it will be popular? Wait until people have to enrol on to the system—they will have to travel to one of the centres, queue up, have their data taken, and be interviewed. And what happens if the scanning goes wrong?
This Bill should be killed at birth. If colleagues think that that is being disloyal to the Government, it is not—it would be doing them a service, not the reverse. It is the most serious piece of legislation that I have ever had to vote on in this House—more serious than the decision to go to war. I urge Members to think clearly about the implications when they decide how to vote.