The functioning of PDS in Indian states of Bihar, West Bengal, Chattisgarh and Odisha has improved owing to the switch from paper ration cards to digital ones. The digital ration cards enabled the reduction of grain leakage, thereby increasing transparency and accountability in the disbursement of benefits at the transaction level. However, with the setting up of the UIDAI, Aadhaar replaced other reasonably successful methods of digital identification such as an electronic chip card in Chhattisgarh, a coupon system post biometric verification in Gujarat and another biometric system in Karnataka. Scholars have criticized Aadhaar for implementation inefficiencies such as transaction failure owing to machine malfunction or poor Internet connectivity and increased corruption in terms of quantity and quality of grains disbursed.

There are two steps involved in the use of digital identity for beneficiary authentication towards their access to benefits: registration, enrollment into the PDS and the transaction step where the beneficiary authenticates their biometrics to buy grains. The authors develop a taxonomy based on three parameters to identify the inefficiencies in the two steps involved in the implementation of digital identification: 1. mode of authentication, which can be biometric or non-biometric, 2. source of authentication, which includes central and local databases containing beneficiary information, and 3. the frequency of authentication that may vary depending on real-time authentication or aggregate number of transactions for each instance of authentication.

The authors identify four types of inefficiencies that consistently occur at the registration and the transaction steps: 1. Classification errors, 2. Identity fraud, 3. Quantity fraud and 4. Beneficiary dissatisfaction.

In terms of the three parameters – mode, source and frequency – at the registration level, biometric mode is less prone to identity fraud but the collection and maintenance of data is expensive. The ability of the Department of Food and Public Distribution (DFPD) to integrate the state’s ration card data with data from other governmental programmes for beneficiary verification digitally can reduce beneficiary dissatisfaction. The use of Aadhaar by DFPD is more cost-effective given the most significant costs of data collection, storage and maintenance is taken up by the UIDAI. The de-duplication of beneficiary records during registration requires the source of authentication to be a central database while frequency is only real-time

At the transaction level, the authors find the non-biometric mode to be more effective given that biometric authentication is prone to Internet and electricity failures, malfunction of ePoS machines and poor quality biometric that may be unreliable. The use of smart cards with PIN and an OTP sent via SMS under the non-biometric mode addresses the authentication concerns of senior citizens with limited mobility and poor quality biometrics. The beneficiary authentication using central database enables beneficiaries to authenticate from any location, however, such source requires robust Internet connectivity. While such an option allows real time monitoring of quantity of grains collected by beneficiary and decreases quantity fraud, the higher probability of transaction failure may lead to higher levels of beneficiary dissatisfaction. The aggregated frequency of authentication, or one-time authentication for an aggregated number of transactions combined with the issue of coupons may increase beneficiary dissatisfaction due to long travels and wait times.

While both methods, biometric and non-biometric offer certain advantages at the registration and transaction levels, Indian states need to quantitatively assess the cost and benefits of each of these choices to customize their use of digital identification including Aadhaar and non-biometric methods. Such studies would enable the states to customize their approach to the particularities of the regional contexts.