EconReads
Donate

Digital Government in India

When Technology Excludes: Authentication Failures

What research found when biometric authentication was introduced for food rations, and why digital systems must be designed to avoid shutting out the poor.

Digital systems can reduce fraud, but they can also exclude genuine beneficiaries.

The Jharkhand study

In 2016, Jharkhand required Aadhaar-based biometric authentication at ration shops. Economists Karthik Muralidharan, Paul Niehaus and Sandip Sukhtankar studied its rollout using a large randomised evaluation, published in 2020.

They found:

  • No significant reduction in leakage overall.
  • Higher transaction costs for beneficiaries: more trips and time spent at ration shops.
  • Some households lost benefits, especially where authentication failed.

Why authentication fails

  • Worn fingerprints of manual workers and the elderly.
  • Poor internet at ration shops.
  • Machine failures.
  • Mismatched records.

Other reports

Media and researchers reported cases of people denied rations or pensions due to authentication problems, raising concern about hunger and exclusion.

Design lessons

  • Fallback options: allowing OTPs, iris scans or manual verification.
  • Offline authentication.
  • Grievance systems to fix errors quickly.
  • Pilot testing before large rollouts.

The broader point

Technology is a tool. Its value depends on design, infrastructure and safeguards. For welfare, avoiding exclusion of the vulnerable matters as much as reducing inclusion of the ineligible.

The failed fingerprint

An elderly woman who spent decades doing manual labour can't get her fingerprints recognised at the ration shop. Without a fallback option, she goes home without grain. A system with OTP or iris backup would have helped.

Thinking more technology always improves welfare

Poorly designed digital systems can raise costs and exclude genuine beneficiaries.

Key takeaways
  • A Jharkhand study found biometric authentication raised costs without significantly cutting leakage.
  • Worn fingerprints, poor connectivity and machine failures cause exclusion.
  • Fallbacks, offline options and grievance systems help.
  • Avoiding exclusion is as important as reducing fraud.
2 min read

No recording for this one yet - EconReader can read it aloud for you.

Welcome to EconReads

This site is made for visually impaired learners, so our read-aloud reader is already switched on to help you explore hands-free.

You're in control - turn it off any time using the Reader button at the top of the page.

EconReader Ready