17 September 2009

What's Green, Biometric and Friendly?

India's Unique Identification Authority (UIDAI), an agency of the national Planning Commission with responsibility for implementing the Multipurpose National Identity Card, has indicated that some 600 million people are to be identified with a biometric card.

That initiative - variously hyped as a solution to government corruption, electoral fraud, illegal migration and terrorism - is estimated to cost around $3 billion and represents one of the world's most ambitious IT projects. Its proponents, undeterred by the collapse of similar super-projects in the UK and elsewhere (including recurrent proposals for a multipurpose Australia Card), envisage the new card scheme as a catalyst for improved delivery of public services and as something that will drive growth in India's IT sector. Perhaps unsurprisingly, given governmental infatuation with the IT industry, the UIDAI is headed by Nandan Nilekani, a software mogul who announced that
We are going to have to build something on the scale of Google but it will change the country … every person for first time [will] be able to prove who he or she was.
Being able to identify who he or she is, rather than was, is of course useful. The shift towards the dizzy heights of e-government via the Planning Commission's latest Five Year Plan might be facilitated if the Commission updated its online media releases (the latest of which dates from early 2007). Finance Minister Pranab Mukherjee stated in July that establishment of the UIDAI is a major step in improving governance regarding public services and
marks the beginning of an era where top private sector talent in India steps forward to take the responsibility for implementing the projects of vital national importance.
Nilekani is reported in today's Guardian as indicating that the scheme will be particularly useful for India's poor, several million of whom this blogger notes are currently reliant on dried cow dung for heating and cooking, live in unsewered houses (if not in the street) and are illiterate.
This will mean maids and labourers … a hundred or two hundred million people – will be able to access welfare benefits for the first time without any questioning who they are.
They will presumably be reassured that the card will not feature caste or religious identification, although in the absence of meaningful privacy law (statutes and enforcement) there would appear to be scope for data matching as individual agencies deploy information associated with the cards. There are suggestions that much of the data storage/handling wil be outsourced. In August the Prime Minister's Office announced that the UIDAI Council had "resolved to provide a legal framework to the UIDAI in the course of time".

Nilekani explained that
We are not profiling a billion people. This will provide an ID database which government can access online. There will be checks and balances to protect identities.
But wait, there's more. He's reported as being in talks to "create a personalised carbon account so that all Indians might buy 'green technologies' using a government subsidy.

Improved service delivery is an excellent thing. Hype about national ID cards as a low-cost low-risk 'silver bullet' solution to diverse problems is however worrying. It is an echo of past enthusiasm from journalists, IT boosters and politicians for the Indian Simputer, a device that would supposedly revolutionise Indian life and fuel the growth of the nation's IT industry. Alas, bold forecasts about the Simputer (and an even bolder US$10 version of the OLPC) have not come to fruition.

Nilekani has elsewhere indicated that
This project is pro-poor and inclusive targeted mainly towards the poor. The middle class and the rich have some form of identity. People on the margins are getting lost because of lack of identity.
The UIDAI thus appears to be aiming to draw together data regarding the identification of some 80 million income tax payers with a Permanent Account Number (PAN), over 200 million account holders in public and private banks, some 500 million mobile phone consumers, over 600 million Indians who have election cards, Indian passport holders, along with databases (often manual) covering such things as cooking gas consumers and holders of public food distribution system cards.

Nilekani admits that validation of data is a challenge, given the "phenomenon of duplicates or fraudulent numbers" and the absence of identity documents for many people (who are for example not individually identified in official registers, don't have accounts with financial institutions and indeed may not have any papers whatsoever).

Wariness might be heightened by a lack of information about what is planned and how the scheme is supposed to work, with some reportage claiming that people will receive a unique number (so much for chatter about photo and fingerprint biometrics)rather than a card, with that number appearing on a plethora of identity documents such as election identity cards, PAN cards and bank account registers. That sounds like an Indian version of the US Social Security Number.