With rigorous economic research and practical policy solutions, we focus on the issues and institutions that are critical to global development. Explore our core themes and topics to learn more about our work.
In timely and incisive analysis, our experts parse the latest development news and devise practical solutions to new and emerging challenges. Our events convene the top thinkers and doers in global development.
Using Biometrics in Development: Lessons and Challenges (The Guardian)
February 11, 2013
Senior Fellow Allan Gelb and Policy Analyst Julia Clark write a piece for The Guardian on how identifying people based on physical traits could benefit aid work, but caution over issues of privacy and fraud.
Citizens of rich countries take official identification for granted. But many in poor countries lack robust IDs, or indeed any documentation at all. This "identity gap" has been an obstacle to inclusive development in many countries. But increasingly, governments and donors have turned to digital fingerprints, iris scans, and other biometrics to provide inclusive, secure and accurate identification for their citizens, from national IDs, to elections and social welfare payments. In a recent Center for Global Development working paper , we surveyed 160 cases where biometric identification had been used for such programmes in over 70 developing countries — cases which cover over 1 billion people!
Unlike paper-based systems, biometrics can reliably ensure uniqueness (have you already registered to vote?), authenticate transactions (are you the owner of this debit card?), and help create an auditable trail (did you already receive payment this month?).
Thus, as mobile technology has done for communication, biometric identification may allow developing countries to leapfrog past traditional paper-based identity systems.
But despite this potential, the technology is no guarantee of success. Concerns remain regarding privacy (though biometrics can improve anonymity when used in place of personal information like name, gender or address), cost and implementation. To date, there has been little empirical work on the effectiveness of biometrics for developmental purposes. Companies with proprietary technology have been reluctant to provide performance data. India's open standards may change this; its performance reports provide a benchmark for accuracy and inclusion that can be used by other countries. India's Unique ID programme now covers some 240 million people, and has contributed to the rapidly falling costs of the new technology. But many programmes are still in the early stages, and only a few of the 160 cases have had impact studies so far.
Still, it is possible to distil a number of important lessons and pitfalls from the information that is publicly available for existing biometrics programmes. In our paper, we identify important successes, risks and failures, and discuss the important role that donors — who already fund a majority of the surveyed cases — can play in identification strategies.
Examples of successes:
Biometrics has helped to promote development in certain cases, particularly with regards to increasing efficiency, improving service delivery and accountability, and allowing poorer countries to build sophisticated identification systems in a relatively short period of time. Biometrics can also facilitate inclusion for marginalised groups; a robust identifier can replace the need for extensive documentation (which the poor often lack) in many cases.
• In Nigeria, introducing biometrics into the federal pension system eliminated nearly 40% of the beneficiary roll. This not only increases efficiency and accountability, but has the potential to improve services (assuming the funds saved are redirected to other services).
• Bolivia was able to expand its voter roll to large section of the population previously absent from the political process.
• South Africa has been using biometric identification and electronic transfers and ATMs to distribute pensions and social grants for over 20 years.
• In Pakistan, an efficient national identification agency (NADRA) ensured transparent management of disaster relief funds by ensuring that only those affected by the 2010 floods received reconstruction payments through Visa cards.
Risks and challenges:
In certain cases, projects have failed to deliver due to problems with logistics, incentives and sustainability. In other countries, multiple agencies or donors control inefficient, incompatible systems that can't be integrated or scaled up. People who cannot provide biometrics (eg, those with damaged or missing fingers) may find it difficult to register, and others may be excluded if identification exercises define citizenship in an exclusionary way. Finally, ongoing concerns about privacy remain — though many cases indicate these are less than often assumed — particularly in countries with weak legal safeguards.
• Voter registration, for example, is a high-stakes activity that must often be done quickly. Though biometrics may seem ideal for providing clean electoral rolls, the hurried timeline can complicate procurement (as recently seen in Kenya) and compromise quality.
• A number of countries have a multitude of IDs, which overlap and cause confusion and waste. In 2006, Nigeria established a committee on harmonisation of national identity cards to address this issue. The committee identified 12 ongoing ID projects at the federal level, eight of which included biometrics.
• While completing a national identification project to include the undocumented in social protection programmes, the Dominican Republic simultaneously implemented laws that effectively stripped many individuals of Haitian descent of de facto, if not de jure, citizenship. This may be a concern in other countries with migratory populations.
Role of donors:
Donors have played a critical role in the rollout of identity programmes and technology, supporting at least half of the surveyed projects. Though critical, this support is often given without long-term strategies for identity management, resulting in fragmented or duplicate IDs. Beyond continuing to play an important demand-side role in the provision of identification, donors can also help set technology standards and bring together disparate parties and agencies to avoid fragmentation and create economies of scope and scale. Because biometrics make it easy to verify the number of unique enrollments, donors may also consider using output-based financing for identification.
Ultimately, there is no perfect system or approach to identification. What works in one country may not work in others, and what works in one local project may not work on a larger scale. Some countries (and regions ) have favoured a supply-driven approach, focusing on providing "foundational" identification (like civil registries and national IDs ) that can be used for many purposes. Others have strengthened identification for a particular service (like elections or social transfers) in response to demand. Like the US social security number, these "functional" IDs may later be expanded for other purposes. Both pathways offer advantages and disadvantages — but in both, incentives and capacity will dictate success or failure.
• For a fuller discussion of these cases, lessons and issues, see our paper .