From Biometrics to Chatbots: Two Technology Challenges Aim to Help Governments Digitize their Economies
There is growing consensus that emerging-market governments should digitize their economies, starting with the vast streams of payments they themselves make and receive. The DFS Lab is launching two technology challenges to fund new technologies that enable this transition: our Biometrics Challenge and our Chatbots Challenge.
The case for digitization
The McKinsey Global Institute tells us in a recent report that “widespread adoption and use of digital finance could increase the GDPs of all emerging economies by 6 percent, or a total of $3.7 trillion, by 2025.” How McKinsey arrives at these magic numbers is always somewhat opaque but whatever the exact figure, it’s clear there is huge value to digitization.
A more nuanced study from the Center for Global Development by Dan Radcliffe shows the policy benefits of government-to-person payment linkages include redirecting energy or food subsidies from producers to individual consumers, and redistributing tax revenues to garner support for things like carbon taxes. The study (citing Global Findex data) shows that there are at least 150 million people in the developing world receiving cash payments from their governments – a terribly inefficient (and often corrupt) system for distributing money.
How should governments digitize?
The Better than Cash Alliance (BTCA) recently developed a set of guidelines for policymakers after reviewing 25 country case studies; they list 10 key accelerators to digitization. As an early-stage, emerging-market fintech accelerator, the DFS Lab team sees payments digitization as a core enabler of our mission to bring adaptive, flexible, affordable financial services that can change people’s lives. Two areas from the BTCA report stood out as being especially susceptible to technological innovation: developing a unique identification system and digitizing government benefit programs.
Converting identity from a hardware to a software problem
Government benefit programs need to be able to identify recipients to distribute payments to the correct beneficiary. Access to these programs can be most beneficial for low-income people; however, they are the ones who most often lack valid identity documents – especially in developing countries. Further, the lack of dependable ID infrastructure creates an opening for fraudulent financial activities, money laundering and the diversion of government resources.
New opportunities to leverage digital identity solutions in developing countries are rapidly becoming commonplace – chief among them national biometric databases such as Aadhaar in India or NADRA in Pakistan. However, biometric approaches require widespread distribution and maintenance of costly biometric readers and hardware which can be challenging in low-resource environments.
DFS Lab’s new initiative DFS Tech is targeting this constraint by launching an RFP to support new fingerprint capture technologies which are purely software based, robust to spoofing and can capture and verify fingerprints using only the sensors (i.e. camera) on an unmodified Android smartphone. The goal is to facilitate a very low-cost and easy-to-deploy digital solution that can be integrated into apps on existing phones all over the world.
Intelligent bots as digital advocates
Low-income people are often challenged to deal with complex formal systems dominated by written rules, complex documents, formal (vs. colloquial) language, non-local language and cultural norms that disadvantage them. Yet they often need to negotiate these systems to get access to essential financial products and services, assert their rights or claim critical government benefits.
In a recent analysis of cross-country data, CGAP finds seven key financial risks experienced by consumers. Four of these represent challenges clients experience in navigating the system: complex and confusing user interfaces; poor consumer recourse; non-transparent fees and terms; and fraud perpetrated by provider and government staff.
The report highlights several areas where these challenges frequently emerge and singles out government-to-person (G2P) programs as particularly fraught:
“CGAP’s 2014 study of electronic G2P payments in low-income countries revealed recourse mechanisms as a particular weak spot (Zimmerman et al. 2014). Recourse and support options were often unclear to recipients, making it difficult to solve problems or get answers to questions they had about their payments. G2P recipients also worried that if they complained they could lose their transfers, a misperception that made them reluctant to report problems. These difficulties undermined the financial inclusion and efficiency objectives of using e-payments for the schemes.”
A particularly deep area of need are interfaces to government subsidy and social safety net programs.
DFS Tech’s second challenge is to engineer innovative natural language interfaces (chatbots) to act as guides, advisers and advocates for the millions of low-income people around the world who interact with these systems every day. These bots would help people understand the program and navigate them through challenges such as how to sign up, add beneficiaries, change address, petition for missed payments and many other important issues. A great analogy for what is possible in this area is Do Not Pay’s “Robot Lawyer”, which helps people in the developing world fight parking tickets and landlord abuse, and file asylum claims – all through a simple chatbot interface. They have gotten more than 200,000 tickets overturned by formulating a legal strategy in the form of a letter to the parking authority.
We have already been in touch with some teams applying to these challenges and are excited to see the technologies they create. If you want to take a shot at applying for either one, or learn more about DFS Lab and DFS Tech, click here and fill out a short application.
Jake Kendall is the director of the Digital Financial Services Lab.
Photo by jerjones, via Flickr.
Homepage photo by William Clifford, via Flickr.
- Categories
- Technology