Wednesday 22 May 2013

Biometrics: a response from the Center for Global Development

Biometrics: will the Center for Global Development reconsider? was published on this blog 10 days ago on 12 May 2013.

A response from the Center for Global Development has now kindly been sent.

On the principle of equal prominence, their response is reproduced here:
Alan Gelb said...

We agree with a number of points raised by David Moss. One is the importance of releasing field performance data; other programs should be held to this standard. We recognize that biometrics is not a panacea. Our previous working paper that reviewed some 160 cases noted several problematic examples, particularly in the area of elections. It is far too early to assess the UID program record in delivering more effective and inclusive services. Where we differ from Moss is that we see the data that it has released on inclusion and accuracy as a very significant benchmark for biometric systems in developing countries, and a major advance on the use of laboratory data. These appear to be the most extensive field data released so far.

The UID data are of interest for other countries; the hypothetical example of Ughana illustrates what such a system should be able to achieve for a “typical” country with about 30 million people. It is easy to scale the results for country size. We estimated that for a country as large as India there would be somewhat over 3 million false positives during enrolment, a large number for manual follow-up but probably doable. For a small country like Haiti the number would only be around 300.

On multiple identities, no system will be able to guarantee 100 percent accuracy. Certainly not the systems in place in the rich countries where identity theft is hardly unknown! The question is not “whether it works or not” but the precision of one system versus another and relative cost-effectiveness. For some applications, such as access to a health insurance program, one might accept a modest level of duplicate or false identities. For others, such as access to a nuclear facility, we want to minimize them – just as we would want very high standards for aeroplane safety, to take the example cited by Moss. These might involve different biometrics and also passwords or other identifiers; the most demanding applications can apply whatever other additional checks they choose outside the scope of national identification. For a national ID system the reported rate of 0.035 percent for UID seems low enough to discourage most deliberate efforts to acquire multiple identities.

Any identification system will have to cope with people who are unable to enroll using biometrics and with identification and authentication errors. The UID data offer useful pointers to likely numbers.

UID does not, therefore, provide answers to every question -- it is far too early for that and we do not claim that it does. It remains to be seen how the program is or is not picked up by various applications and how it negotiates the political winds that arise with any system of identification. But we hold to our conclusion that the data released provides a very significant benchmark on the capabilities of biometric systems in developing country conditions and one that should be studied carefully by other countries.

To correct the record, we do not assert that the number of bilateral comparisons is the square of the population, n. It is 0.5*n*(n-1) which rises (as we note) with the square of n. As n becomes large, it approaches 0.5*n*n; since no identification system will cover 100% of population, we rounded n off to 1 billion for India. If we accept the field estimate of 0.057% false positive rate against a data base of 84 million, the rate for a 1:1 comparison would have to be very small, in the range of 7 in one trillion. The implied precision can only be possible with the combined use of multiple biometrics, which is another of the lessons from the UID exercise.

Alan Gelb,
Senior Fellow,
Center for Global Development

21 May 2013 22:17

No comments:

Post a Comment