Digital public infrastructure and the digital divide

The first principle

Roads are built by the state, but trucks, buses and shops are run by everyone else; nobody expects the government to own every vehicle. Digital public infrastructure (DPI) applies the same idea to the digital economy: the state builds a few open, shared rails for identity, payments and data, and private and public innovators build services on top of them. The rails make services cheap and universal, but a rail helps only those who can reach it, so the same note must also deal with the people the rails do not yet reach, which is the digital divide.

The core

What makes infrastructure public

• Open and interoperable. DPI is built on open standards and application programming interfaces, so any bank, app or department can connect, and a user can move from one provider to another without losing access.
• Minimal and reusable. Each layer does one thing, such as proving who you are or moving money, and does it for every sector, so a new service does not need its own identity system or payment network.
• Public rails, private innovation. Governance of the rails stays with a public or not for profit body, while competition happens in the apps on top. This is the opposite of a closed platform, where one company owns both the rail and the market.
• Safeguards in the design. Consent, data minimisation and grievance redress have to be built into the architecture, because a shared rail failing or being misused affects everyone at once.

The India Stack

• Identity: Aadhaar. The Unique Identification Authority of India (UIDAI), set up in 2009, issues a twelve digit number linked to biometrics. The Aadhaar (Targeted Delivery of Financial and Other Subsidies, Benefits and Services) Act, 2016 gave it a statutory basis and allows its use for authentication in welfare paid from the Consolidated Fund of India.
• Payments: UPI. The Unified Payments Interface (UPI), launched in 2016 by the National Payments Corporation of India (NPCI), lets any bank account send money to any other through a phone in real time, with many competing apps on one shared network.
• Documents and data: DigiLocker and Account Aggregators. DigiLocker (2015) stores government issued documents such as marksheets and driving licences in verified digital form. The Account Aggregator framework, created by a Reserve Bank of India (RBI) direction in 2016 and live from 2021, lets a person share financial data from one institution to another only with consent, in line with NITI Aayog's Data Empowerment and Protection Architecture (DEPA, 2020).
• Sector rails. CoWIN (2021) handled registration and certificates for COVID 19 vaccination; the Ayushman Bharat Digital Mission (2021) gives health IDs and links health records; and the Open Network for Digital Commerce (ONDC, 2022) lets buyers and sellers on different apps trade with each other.

Dimensions

• Economic. UPI made small digital payments almost free, which brought street vendors and small shops into the formal payment system, and Account Aggregators allow loans based on cash flow rather than collateral.
• Governance. Aadhaar authentication and bank transfers made Direct Benefit Transfer (DBT) possible at scale, removing duplicate and ghost beneficiaries; that story is told in the note on DBT and the JAM trinity (Jan Dhan, Aadhaar, Mobile).
• Legal and rights. In Justice K. S. Puttaswamy v Union of India (Aadhaar case, 2018), the Supreme Court by four to one upheld Aadhaar for welfare and subsidies under Section 7, but struck down Section 57, which had allowed private entities to demand Aadhaar, and rejected compulsory linking with bank accounts and mobile numbers. Privacy as a fundamental right was already settled by the nine judge bench in 2017. Data protection now rests on the Digital Personal Data Protection (DPDP) Act, 2023, covered in its own note.
• International. The G20 New Delhi Leaders' Declaration (September 2023) recognised DPI as a route to inclusive development, welcomed a G20 framework for DPI systems and India's plan for a Global DPI Repository where countries can share their solutions.
• Social. Shared rails lower costs for everyone, but authentication by fingerprint and app only processes can shut out the elderly, manual workers with worn fingerprints and people without phones, so design decides who is included.

Source: National Payments Corporation of India, UPI product statistics.

The digital divide

• Access. Many villages still lack reliable broadband, and households often share one phone. The National Statistical Office survey on education (2017 to 18) found that about 24 per cent of households had internet access, about 15 per cent in rural areas against about 42 per cent in urban areas.
• Ability. Owning a phone is not the same as being able to fill a form, read a one time password or spot a fraud call; digital literacy is the skill gap behind the access gap.
• Affordability. Data in India is cheap, but a smartphone and the cost of travel to a service centre still weigh on poor households.
• Language. Most government portals work best in English or Hindi, which excludes speakers of other languages.
• Gender. The National Family Health Survey 5 (NFHS 5, 2019 to 2021) found that about 33 per cent of women had ever used the internet, against about 57 per cent of men.
• Rural and urban, and disability. Rural households, older people and persons with disabilities face all of these gaps together, and inaccessible websites add another barrier.

In practice

Has the divide held back rural development?

• Where it has. When ration and pension systems made biometric authentication compulsory, failures in fingerprints, networks or seeding led to denial of rations, and reports from Jharkhand in 2017 linked deaths from hunger to cancelled ration cards. During the school closures of COVID 19 in 2020 and 2021, children without devices or connectivity lost much more learning than those who could study online. Online only applications for scholarships, crop insurance or jobs push rural applicants towards paid middlemen.
• Where it has not. UPI and DBT reached rural users quickly, because they worked on ordinary phones and through bank agents. Common Services Centres (CSCs), of which there are more than five lakh, let village entrepreneurs fill forms and print certificates for people who cannot do it themselves. Assisted models show that the divide in skills can be bridged by a helper even before the divide in access is closed.
• The judgement. Digital illiteracy and poor access have hindered development where systems were built online only and without fallback, but where design assumed low skills and offered assisted access, digital rails have widened rather than narrowed inclusion.

Measures already in place

• BharatNet. Approved in 2011 as the National Optical Fibre Network and renamed BharatNet, it aims to connect all of about 2.5 lakh gram panchayats with optical fibre under the Department of Telecommunications.
• Pradhan Mantri Gramin Digital Saksharta Abhiyan (PMGDISHA, 2017). It set out to make one person in each of six crore rural households digitally literate.
• Bhashini (2022). Under the National Language Translation Mission of the Ministry of Electronics and Information Technology (MeitY), it builds open translation and speech tools in Indian languages so that services can work in the user's own language.

Issues and way forward

• Exclusion by design. Authentication failures fall hardest on the poorest, so every service needs an exception route: a manual override, an alternative identity check or a one time password, with no denial of a right because a machine failed.
• Offline fallback. Services that matter for survival, such as rations and pensions, should keep an offline route and receipts until connectivity and skills are near universal.
• Assisted digital access. CSCs, post offices and frontline workers should be paid and trained to help, because the helper is part of the infrastructure.
• Language and accessibility. Interfaces should be built in Indian languages through tools such as Bhashini and meet accessibility standards for persons with disabilities.
• Governance of the rails. Bodies that run DPI need clear legal mandates, independent audit and grievance redress, so that public rails stay accountable as they grow.
• Women's access. Closing the gender gap needs targeted literacy drives and shared community devices, not only cheaper data.

Common confusions

• DPI and e governance. e governance is the use of Information and Communication Technology (ICT) by the state to govern; DPI is a set of shared rails that the state, businesses and citizens all build on. DPI enables e governance but also private markets.
• Aadhaar number and authentication. Holding an Aadhaar number does not mean one can always authenticate; exclusion usually comes from failed authentication, not from lacking the number.
• Digital divide and digital illiteracy. The divide includes access, cost, language and gender; illiteracy is only the skills part of it.
• Puttaswamy 2017 and Puttaswamy 2018. The 2017 nine judge bench declared privacy a fundamental right; the 2018 five judge bench applied that right to Aadhaar and upheld it with limits.

For the exam

• Keyword bank: digital public infrastructure, India Stack, open and interoperable rails, Aadhaar, UPI, Account Aggregator, consent architecture, ONDC, Puttaswamy 2018, Global DPI Repository, digital divide, digital literacy, assisted access, offline fallback, BharatNet.
• Practise the previous year questions on this topic →
• Use in an answer:
  • Introduction with data: "According to NFHS 5 (2019 to 2021), about 33 per cent of Indian women had ever used the internet against about 57 per cent of men, a reminder that digital rails reach only those who can use them."
  • Introduction with a declaration: "The G20 New Delhi Leaders' Declaration (2023) recognised digital public infrastructure, open public rails with private innovation on top, as a route to inclusive development."
  • Conclusion: "Digital public infrastructure has made services cheap and universal in design; making them universal in practice needs assisted access, offline fallback and interfaces in every Indian language."
  • Conclusion: "The test of a digital state is not how many transactions it records but whether its poorest citizen can still claim a right on the day the network fails."