Winning the DPI race

India is leading the race to virtually create another continent—the eighth—which will be alluring both in terms of functionality and incentives for global citizens to partake of it for their own individual benefits. The eighth continent rests on a foundational digital public infrastructure (DPI), which has set off a race among nations wanting to be part of it. DPI is the conductor in this orchestra and intercedes in the flow of people, money, information, etc, and enables a whole slew of other blocks to be built atop it.

Infrastructure—both physical and digital—gets built across the lifetimes of many governments and cuts across the usual five-year electoral lifecycles. India Stack, the foundation of which was Aadhaar, was launched in 2009 but, 14 years later, it is not only India’s foremost public good but also is becoming a global public good. Many global leaders have sought it from India, and prime minister Narendra Modi has authorised the sharing of its design, framework, and implementation in a statesman-like gesture. India could have held onto it or patented it, and made it harder to share—or even used it as a bargaining chip—but instead it became a shining example of India’s gifts to the world.

Also Read

Stick to the knitting

Data Drive: Microfinance gains

Disclosures and disconnects

Plugging a loophole

The India Stack—a large part of what constitutes the DPI—is unique, given it is a rare collective action of governments (the Union and the states), the corporate sector, a large pool of anonymous but extreme public-spirited volunteers and not-for-profits. It is a rare example of private provisioning of public infrastructure—completely free of cost to citizens and almost free to corporates. Most important, it is at the population-scale and is designed to serve the myriad, dynamic needs of 1.4 billion Indians. 

It is extremely difficult for engendering collective action in even groups of 100 or 200, especially when they represent multiple, often conflicting, interests. The beauty of India Stack is that it brought together disparate interests for a common objective that has been woven together through simple interoperable technologies called APIs (application programming interfaces). APIs, simply put, are a set of protocols enabling two pieces of software in different systems to talk to each other. 

Big Tech is on the back-foot globally and under serious regulatory fire everywhere for sins of omission and commission. At such a time, whether it is DPI-identity via Aadhaar, including eKYC authentication, or UPI (Unified Payments Interface), or a mechanism to share personal data without compromising privacy (the Account Aggregator protocol built on the Data Empowerment Protection Architecture or DEPA), Indian tech is consolidating presence. Other than the examples mentioned, there is DigiLocker, eSign, GSTN, DBT, Fast Tag. Should these be privately provisioned or publicly provisioned or be offered by a combination of the two models? There are examples of each around the world. 

In India, there is often outrage whenever the government collects data for any public purpose, but none when any private company does so! It is no secret that almost every major private corporate has 800-1,000 data points about every one of us based on our shopping, items viewed or purchased, domestic or international travels, journeys undertaken, payments history, etc. This begs the question—beyond the scope of this piece—between a public versus a private monopoly, which is the lesser evil? 

India’s business backbone is increasingly powered by DPI, which, in turn, is powered by APIs; instead of going into their descriptions, this piece highlights just three of the many possibilities and issues around them that will determine their shape and form going forward.

First, UPI seems to be on a rocket-ship. Today, ~300 million transactions, valued at approximately `60,000 crore per day, are carried out UPI. But, imagine if the 1-billion adult population of India uses UPI to transact just 3-4 times a day, this number would rise 10X if not more. What is holding it back? As of now, UPI was designed such that it will always remain free of cost to customers who use it. However, there is a case for the zero-MDR policy to be revisited for large merchants or for large transactions. Given the availability of data on both, it would not be very hard to work out a mechanism wherein, for every transaction, large merchants are charged 20-25 basis points, which, in turn, is plowed back into the system to enhance and make the infrastructure more robust. Payment systems around the world charge some amount for their services and UPI disrupted all by offering zero MDR. Now that it is established and entrenched in the system, it should continue to be free to customers; however, for large merchants, it should be able to charge more than what it does today. This transaction fees can be capped by the government for large merchants. In turn, it should be mandated that the ecosystem builders such as Google Pay, PhonePe, PayTM, and a slew of others that get this fee plow at least some of it back into creating awareness of UPI, especially in smaller towns where it needs to spread further than where it is today. B2B UPI is almost non-existent, and a concerted effort should be made to create and grow that ecosystem also. For all this, the global and Indian fintech companies need a few billion dollars every year, and that can only be possible by levying a small fee on large transactions with a clear, monitorable mandate and built-in incentives enabling the large fintechs to plow them back into the system.

Second, the ONDC (Open Network for Digital Commerce), the next iteration of DPI, unbundles buyers and sellers and reduces the dominance of the e-commerce behemoths. But, the latter can potentially leverage ONDC to extend their reach, though the ad-tech driven business models that play up certain goods won’t work well then. The proposed consumer protection rules appear to be too broad on two elements. If ONDC comes under the definition of e-commerce entity, then its compliance costs will increase and negatively impact the small seller or startups wishing to use the ONDC. Further how, how much—and whether at all—fallback liability should be there, needs further debate. Thus, the applicability of these rules to ONDC should be limited for; else, you risk killing the golden goose! 

Third, India needs to take the lead via DPI to write the global playbook, protocols, and standards for countries to see it as a viable alternative to other models which might fulfil similar functionalities. This can be done through the G20 and other forums, and, invariably there will be a deep variance with others. Thus, DPI must demonstrate superior functionality, visceral simplicity, reasonable affordability in design/deployment and genuine interoperability to win this race.

Disclaimer: Author is an IAS officer. Views are personal.