[Dave Birch] In “Banking the World — Empirical Foundations of Financial Inclusion” by Cull et al, there is a very interesting case study on the use of fingerprinting to manage credit and repayments amongst farmers in Malawi. In Chapter 13, “Use of Biometric Technology in Developing Countries” by X. Gine, J. Goldberg, S. Sankaranarayanan, P. Sheerin and D. Yang, the authors describe how biometric technology worked in practice in a field experiment amongst paprika farmers who had applied for an agricultural loan from a government lender. The farmers were randomly allocated between a control group and an experimental group who had a fingerprint collected as part of the loan application process. Fingerprint recognition was chosen because it was cost-effective and reasonably robust in the circumstances. Despite the farmers having worn and damaged fingerprints, as you might imagine, only 2% failed to enrol with the right thumb and had to use another finger instead.

The study showed that fingerprinting led to increases in the loan repayment rate of almost half amongst the highest risk groups, but had no impact on the repayment rate amongst the low-risk groups. What an interesting result. The authors cite a rough cost-benefit analysis of the experiment which suggested that the benefits of improved repayments greatly outweighed the costs of equipment and fingerprint collection. In both the high- and low-risk groups though, the farmers now had a biometrically-verifiable credit history. Having a basic credit history is, as we all know, are really important step on the ladder of financial inclusion with very beneficial effects.

The authors also noted, however, that fingerprints aren’t tremendously accurate. If you’re going to use fingerprints to identify people from a large population this could be a big problem. On the other hand if you are authenticating people against a card (or a phone, for example) from a smaller population such as farmers who have obtained agricultural input loans in a particular district, then the speed and convenience of the fingerprint authentication make it a reasonable choice. Once you start messing around with fingerprints at population scale, then the cost and complexity rides astronomically.

Police zeroed in on Afsar after UIDAI confirmed that all the authorization fingerprints used for the enrollment of the 60 persons in biometric exception category were his. Afsar, who used to work as a data entry operator with IL&FS, had quit the job in August 2011.

[From Seven booked in Aadhaar fraud – Times Of India]

What I’m saying here is that fingerprints for authentication, in context as a convenience technology, provide what seems to be a much better cost-benefit balance than fingerprints for identification, in context as a security technology. The convenience point is at the heart of the mass market proposition, and it is undoubtedly why Apple decided to take a look at the technology*.

“Apple is not going to the trouble of adding a biometric sensor just so that you don’t have to use a four digit password. They are adding a biometric sensor so that the iPhone can become a safe and secure payment device,”

[From Some Ideas for What Apple Could do With Its AuthenTec Purchase – Technology – The Atlantic Wire]

No, they are adding a biometric sensor so that the iPhone can become a safe and secure identity management device, and one kind of identity it will manage will be financial, and one kind of financial identity will be payments. Apple understands the location of biometrics in the consumer space: convenience, and Apple is all about convenience. Remember, these iPhones aren’t going to be used to launch nuclear missiles or identity people in databases: these iPhones are going to be used to 1-1 local matching of fingerprints to stored templates to authenticate amongst persona. I think one of the first blog posts I ever wrote was about this!

* A couple of weeks before this Apple announcement, which I knew nothing about, I was interviewed for a television programme and I confidently predicted that fingerprint authentication would make its way into the next generation of mobile phones and — for reasons too boring to go into here — specifically mentioned Authentec as supplier of decent kit in the fingerprint authentication field. So I got to big up Authentec on a television programme a couple of weeks before Apple bought them for a few hundred million dollars and made me look like a guru. Thanks Apple!

These are personal opinions and should not be misunderstood as representing the opinions of 
Consult Hyperion or any of its clients or suppliers

1 comment

  1. Your biometric tic is taking longer to treat than perhaps either of us imagined when these sessions started.

    “Apple shareholders turn on guru who advised the company to spend hundreds of millions on Authentec” – ready for that headline?

    1:1 flat print fingerprinting usually produces about 20 percent false non-matches. That’s with the threshold set to achieve as near as possible 0% false matches. Where was the threshold set for the Malawian trial?

    The belief in magic is understandable but, after McCormick, not recommended.

Leave a Reply


Subscribe to our newsletter

You have successfully subscribed to the newsletter

There was an error while trying to send your request. Please try again.

By accepting the Terms, you consent to Consult Hyperion communicating with you regarding our events, reports and services through our regular newsletter. You can unsubscribe anytime through our newsletters or by emailing us.
%d bloggers like this: