The Identity Initiative at ISB

This article provides insight into the Identity Initiative, a unique collaboration between academic scholars, industry leaders and government officials interested in studying the impact of Unique Identification (UID) in the coming decade.

In September 2010, Indian government embarked on an ambitious project to provide a biometric-based unique digital identity, Aadhaar, to each of its 1.2 billion citizens. This UID (Unique Identification) initiative is unprecedented in scale – collaboration of the public and private sectors in voluntary enrolment of citizens – and the openness of the technology platform towards facilitating its use by businesses as well as by government. While weak identity systems cause the individual to miss out on benefits and services, it also makes it difficult for the government to account for money and resource flows across a country. The need for establishing a strong nationally portable identity for its residents was succinctly captured by Nandan Nilekani, Chairman UIDAI, when he stated, “There are just 50 million Indians who have a passport, that’s 5 % of the population. People who participate in the stock markets must be 2% of the population, bank account holders must be 20% of the population (220 million). If you want to meet all the people’s aspirations, you have to include them in the society. To do that, the basic foundation block is identity.”

On the day of the official rollout of the UID by Prime Minister Manmohan Singh and Ms. Sonia Gandhi, we wrote in the Wall Street Journal India (the Mint) about how the true benefits of the initiative will be in reshaping India’s economic institutions.  This led to the formation of the Identity Initiative (IDI), a unique collaboration between academic scholars, industry leaders and government officials interested in studying the impact of this nationwide effort in the coming decade. The IDI will conduct and sponsor academic research, facilitate industry events to seed public discourse, and provide an objective, rigorous view of the UID’s socio-economic impact.

The activities of the Initiative will help:

Our research is an independent academic inquiry, aided by a memorandum of understanding between the Identity Initiative and the UID Authority of India to facilitate use of UID data, provide appropriate access to the UIDAI leadership, address pressing public priorities and communicate research findings and recommendations to policy makers.  Nilekani sees the engagement with “academic institutes like ISB as very strategic and very much a real time engagement as opposed to a post mortem of what we have done.  We would like the academic community to work with us in real time as we roll it out and conduct all kinds of controlled experiments and other studies to see the impact on the ground and we will use the impact analysis as a feedback loop into our system so that we can design our programs and implementation better.”

If you want to meet all the people’s aspirations, you have to include them in the society. To do that, the basic foundation block is identity.

Current Research Projects:
Longitudinal analysis of socioeconomic impact.
This project will use primary survey-based data to examine the adoption and socioeconomic impact of the UID effort between 2011 and 2015. Our data will be from a nationwide panel comprising 100,000 households in half the districts of India, conducted in collaboration with NCAER (National Council for Applied Economic Research).
The survey was initiated in January 2011. Each individual in the sample will be administered a detailed questionnaire on household finance and demographics, along with specific questions on acquisition and use of UID as well as other forms of identification, capturing baseline enrolment levels, spectrum of usage, perceptions of value and peer enrolments. Our design explicitly accounts for voluntary individual enrollment into the program, and the motivations of for-profit UID registrars that may lead them to offer UID in those areas or to those individuals who may be most profitable. It also considers the dependence of the UID’s effectiveness on the roll out of associated applications in sectors such as public health and education, financial services and social insurance.
A key initial focus of the data analysis will be on whether the UID is altering the welfare of India’s 100 million migrant workers, and if so, what mechanisms are generating this growth in surplus.

Merging market and government facilitation of national infrastructure
The UID Authority has chosen an interesting and complex combination of public and private sector resources in rolling out Aadhaar numbers and to grow the ecosystem around the UID platform. This project uses a mix of methods to analyze different combinations so that we can add to our knowledge about the relative benefits of market vs. state provision of institutional underpinnings.
One part of this project examines the design of the UID Authority’s rollout ecosystem, focusing on its three-layer supply chain, and the different decentralized designs for rollout and marketing operations, while preserving centralized control of authentication. It will also examine the unique combinations of state governments, national banks/insurance providers and outsourcing agencies being used. A part of this project will use natural experiments in Andhra Pradesh to contrast the different operational strategies employed by public and private sector registrars and the associated patterns of enrolment.
A separate part of this project will analyze the openness of the UID platform as well as the partnerships that the UIDAI is forming to “seed” integration of the technology into different financial services, mobile and healthcare sector applications, along with the related approach towards supporting R&D that will enable wider adoption of the low-cost biometric devices essential for the eventual use of Aadhaar numbers by businesses and government.

Using RCTs to measure the value of Aadhaar
Many factors affect the selection of people who adopt the UID, the fraction of these that actually use it, and the benefits that eventually are realized by businesses that integrate it into their products and services. This project will enable a series of controlled experiments that will allow a more rigorous measurement of the value created by UID adoption.
One set of trials will measure the economic impact of UID-enabled applications in partnership with Eko Financial Services, a provider of small-scale banking and remittance services (ATMs with human tellers) to the underprivileged in Bihar and Jharkhand. These will involve defining the set of villages that Eko will launch its UID-enabled services in according to our experimental design and then measuring the impact of their service on the village relative to a matched control set of villages at which the services were not available. A different set of trials will estimate differences in outcomes for those individuals who utilize the UID as compared to those who do not, using a researcher-led intervention that will educate targeted “treatment” groups about the benefits of UID adoption and its associated applications, while observing the behaviours and outcomes for these groups as well as corresponding “control” groups that do not receive this intervention.

We would like the academic community to work with us in real time as we roll it out and conduct studies to see the impact on the ground. We will use the impact analysis as a feedback loop into our system so that we can design our programs and implementation better.

Modeling the diffusion of Aadhaar
The first step towards realizing the benefits of the institutions that will be created or strengthened by the UID is to ensure that adoption is as rapid and widespread as possible. This project uses individual-level data to create a structured meta-population model of the diffusion of Aadhaar numbers, towards understanding the factors that might optimally accelerate its spread.
Many factors could influence successful adoption diffusion, including the level of centralized marketing and communication about the UID; the extent of local (village-level) communication by on-the-ground volunteers; the influence of key village community leaders; spillovers from news of successful enrolments in neighboring villages; efficiency of the enrolment agency that is assigned to that specific region; and the effectiveness of the district-level government leadership.
Structured meta-population models have been used successfully in the last couple of years to explain epidemic spreads such as that of the H1N1 virus that did not fit the patterns predicted by traditional SIR and SIS epidemiology models. Our research adapts these models in the context of rural technology diffusion, while also developing extensions that account for a variety of unobserved covariates.