Latest / Computer Says Maybe / The Vaporstate: ID in India

Transcript

The Vaporstate: ID in India · Computer Says Maybe

• Astha: [00:00:00] DPS like a way of life. You just have to adopt it. Mila: Each person needs a unique id. Each land person needs a unique id. It's super easy. It's for free. Somebody else with Rafa: her identity was opening telephone accounts. Beatriz: Oracle is in so many different government systems. Dan: You build an entire tech stack on big tech products, and then you. Put a union jack on the top of it and you wave it intensely in the hope that nobody will look at what it's sitting on. Alix: This is Vapor State. A new series from us here at Computer says May. So we're gonna start in India, which is home to a pretty well established, you probably have heard of it, program called aha, which began as an attempt to create a unique ID for everyone that needed access to public services in India and to help better govern that program. I use the word govern, which we'll get into a little bit with a couple of our guests that maybe that's not an apt word as to how, [00:01:00] um, the program began because there actually wasn't even a law passed when it was initiated. Despite how sensitive and complex the program was meant to be. In this episode we feature Mila Samdub and Astha Kapoor, who I sat down with the three of us. Um, we talked through, um, their analysis of how Adhar has evolved and the kinda political implications of this digital IT system. And it's a really good grounding, I think, in a lot of the common pitfalls and sort of political dimensions of these kinds of projects. Um, I also spoke with Usha Ramanathan separately, who is. A kind of a legend, um, lawyer in India who has fought all kinds of fights, including spending the last 15 years or so, sort of systematically challenging state expansion of Othar, which you'll hear more about directly from her. She'll pop up throughout with, um, really good legal analysis that. Help me at least kinda wrap my head around the questions I should be asking. If a country near you says, Hey, wouldn't it be great if we had a digital ID system here? I feel [00:02:00] much more prepared and armed with OSHA's powerful questions and legal perspective on the topic. So hopefully you will leave the episode feeling empowered to ask really good questions and sort of see. The political complexities and sort of think through a lot of the stuff yourself, no matter what country you're in. So, without further ado, let's get into it. Astha: I'm Asta Kapu, I'm co-founder app, the institute. We're a research and advisory firm based in Bangalore, and we look at the impact of technology on people. Mila: Hi, I'm Melissa Sub Dubb. I'm a designer, writer, and a curator. I work on the aesthetics and political economy of digital infrastructures. Mostly in India, but also elsewhere in the majority world. Alix: Uh, I'm wondering if both of you could take a crack at a high level definition of DPI, which I feel like I know is like the bane of your existence is when people are like DPI. What's that? You're like, give me an hour. Um, so how would you define it? Astha: You okay. I'll go. [00:03:00] Uh, so I always, I always say DPS like a way of life. It's like Hinduism, you just have to. Adopt it. It's not definitional. It's called an approach. It's called a way of thinking. And the reason for that is that it's not one thing, it's the pantheon. So it means open and interoperable and extensible infrastructure for public interest that is governable. So it's very many things, but I think. When people do say DPI, it's reusable tech at scale. Like that's the easiest way to understand it. But it's not easy, it's complicated and therefore it's an approach, like I said, and it's often called that. So, you know, in the context of DPI in G 20 is the DPI approach to say public tech, but it's digital. Public. Alix: Public nominally. I mean, we can get into whether it's public and it's infrastructure, so you're supposed to scale it. Then it's [00:04:00] a cost effective way to expand access to some type of service that ideally makes people's lives easier. Astha: Yes. In the public interest is one of the definitions of the public in DPI and all of these app photographs like digital, public and infrastructure, you could unbundle them in many different ways. Mila, do you wanna take a Alix: crack? How would you define it? Mila: Yeah, I think, um, more than anything, DPI is a vibe. But we can sort of unpack that vibe and there are obviously parts of the vibe or there are things there that are much more concrete. Everyone is sort of nominally agreed. Countries have nominally agreed on a definition of DPI is interoperable, open, and at scale they've done that. Like at the G 20, the UN digital compact and so on. But that like doesn't tell you a whole lot about what these things look like. And so functionally, apparently it means like digital ID systems, payment systems and data exchange systems. But nobody agreed to that. And you keep getting like this phrase [00:05:00] that, oh, there's a broad consensus that the DPI refers to these three things. So there's like a functional way of understanding like, what are these things actually? And, and that kinda gives a bit more materiality to it maybe. And then I think there are also like architectures that are common between all of these systems. One is that. There's like a mania to assign unique IDs to things and that that's like, that's like really what underlies all GPI like. You know, each person needs a unique id. Each land parcel needs a unique id. Every student needs a unique id and so you get databases with unique IDs and that's not so different from like how surveillance technology works in general. Right. And the other is kind of a platform ecosystem architecture, which is also not so different from how a sort of big tech model of tech works in general, where you have states building platforms or state adjacent organizations building platforms that other sort of state agencies or actors can access. And so maybe from the kind of really broad vibe to the more sort of concrete architecture, I'll also say that maybe like. In practice since DPI is such a [00:06:00] fuzzy label and like it could mean anything if you wanted to give it sort of a more concrete label, it might be something like public private software platforms. Usha: I'm Usha Raman, and I'm based in Delhi, India. I'm quite old, who I've been working for many, many years on issues of law and poverty, the jurisprudence of law, poverty, and rights. And in the past 15 years, thanks to the UID project, I found a new avenue that's entered course Icot was connected with law and poverty, which is technology. In the 1990s, we had the first introduction of the voter ID card. Initially we were all very worded because we were saying that the poor lived in a twilight zone of legality. It's very difficult to be poor and to be legal, so we've always felt that shadows were important, that the state should be transparent, but people should be able to find the few shadows to secure. Sales, tracking every person down and giving them an ID seemed like it [00:07:00] could be a problem for the poor. But actually, we formed in the nineties, that card did not harm them. In fact, it ate them in quite a few ways, including in establishing that they had a right to the city, for instance, because they'd been there since the card had been issued in various things with that process. So when the UID project, the other project was originally spoken about, in fact, it was hardly. Spoken about. There were no big news reports on it. There was hardly anything. Uh, how we got to know about it was because they held meetings with what they call civil society actors to introduce us to the concept, and I think basically to co-opt us into making the project work because we have access to certain kinds of people to whom they don't have necessarily an immediate. For instance, migrant workers, people working with migrant workers, they say, if you can come in, if you can, you know, onboard onto this project, it'll help. So we didn't go into it with skepticism because we thought that maybe this will be another such where it help the poor. But the first meeting that we had told us that. [00:08:00] The ambitions were very different, or you know what the ambitions were, we didn't know. But this much was clear that the people who were working on the project really didn't know the poor at all, and it seemed like they were shooting from the shoulders of the poor. We asked them questions. It received irritation in answer. You have to investigate these things, and the project started without a white paper, blue, blue paper, green paper, whatever color paper you want. So nobody knew where this. Project, what this is supposed to do, what it enticed people to, what are the risks that might be involved in it. But very soon as the documents started emerging, for instance, there was a biometric standard committee which produced a report, which was to say how they're going to adopt biometrics in the project. We realized that this is a project which is being rolled out on a national scale, but without really knowing either the science. Or the sociology of it. And that made us ask many questions. And I think the reason that the skepticism grew is because we haven't received those answers still today, [00:09:00] Astha: post 2014, I think things changed fairly rapidly, which is that the UDI remained in. As a government, I don't even know how it's structured anymore, but it remains a part of the government. But a lot of the implementation, whether it's the enrollment, whether it's the authentication, whether it's the use of the know your customer function, all of that has been over time. Moved more and more to the private sector, which one, one would argue is that the, the ID is public, but the use is private, which is how it was always maybe envisioned. And I think what we are also seeing is that the Supreme Court in the AHA act, which was what 2017 had said that it would be used for certain functions and you may not be able to use for other private functions, such as getting a SIM card or bank account and things like that. But now. Even eight years since that act, we have seen some of those things also getting pulled behind. So the private sector is [00:10:00] now able to use the digital ID for many functions that were not even originally allowed in the act. I wouldn't say that the UIDA is a private thing, but the way that the use has been structured over time, it has become increasingly. Privately used, Mila: there was always a private sector component to the provisioning of AAR or the infrastructure that is within the infrastructure of aar. So, for example, enrollment was always outsourced to private sector firms. When you actually look at how the data flows within aha, you've always had these kind of infrastructural intermediaries that, you know, the end user never sees. And those have always been large commercial entities. And part of the reason those those functions were outsourced was so that they could get scale really quickly. And also so they could sort of bake in private sector interests into the sort of coalition that was building this at the time. But yeah, as I think you're absolutely right in that the use. At the end use of it gets really expanded for [00:11:00] commercial end users after 2014 to the point where the infrastructure is basically agnostic to who's using it. Like a government agency could be requesting your other, your identification using it or. A telecom operator could be requesting your identification using aar, and it's kind of agnostic to whether it is a state agency or a private sector actor that's doing that. Astha: In 2000 9, 10 11, it was always used as a proof of will you being you and a proof of residency in India, right? Like that was the vision of it. So that was the first sort of like, this is the basic thing. It's going to only have your name, your gender. And your age. I think those were the things, because I remember being in the planning commission when it was being demonstrated, and people asked like, how can this be misused? And they were like, oh, if somebody wanted to target all women, only then can this be misused? So, you know, because it was so data light at the time, and then it was like. It's going to be used for deduplication of [00:12:00] government subsidies because there were enormous leakages in India's food, fuel and fertilizer subsidies. So it was going to be used for that. But then as Mila was saying, is that very quickly the use cases started to ramp up to say, this is not just deduplication for. Subsidies of the government of India. This is actually a know your customer function for banks and sim cards. Okay, now we need to connect it with other things. So it's just been this sort of hive mind that's been trickling through various things. And as of what last week, the Supreme Court wants to use it as an authentication mechanisms for people going to. Concerts almost 12 years later. What started off, just as a proof of residence, is now becoming, you know, your entry way into your private life if you're going to concerts or wherever else. Alix: How old were you when Adhar, I don't know, maybe that's a rude thing to ask, but how, how old were you guys when it, when it [00:13:00] went live? I Astha: was 23, 24. Uh, when it started with some pilots in 2011. I encountered it very early because I was in the government at the time, so I encountered it in the botched up pilot phase where, you know, things were going wrong and things like that. The rollout started fully in 2011. 12. Alix: I didn't know you were in gov. So when you were in government, was, what was it presented as? Like, this is gonna make your life easier, or the tech weirdos have some new thing they, they want you to play with? Or like how was it framed? Astha: It was magic. It was like, you will be able to, I authenticate Indians based on biometric identity. Indians either had like a proliferation of ID or didn't have any id. So you know, that was the halves who have very many things. And then there was a huge chunk. You also had use cases like in India, which is like, oh, when women get married, their names get changed, and things like that. So [00:14:00] this is going to tie your identity to yourself. So even when we were in the. Field. We actually found that for women in particular, this was an incredibly empowering thing because it would, you know, take care of loss or change of names and things like that. And so people would say, this is a part of who I am and nobody can take it away from me. And that's how it was sold, which is that. We need this unified ID system in a country as complex, large as India. But again, this was also parallel to another effort which got junked, I think the national population register, which was a citizenship effort. And this was just a residency. So for instance, even now, if you live in India for more than 180 days, you're eligible for an Alix: how exciting for you. Yeah. Okay. So it was like. Basically, let's get organized as a state and be able to make it easier for individuals to interface with us. Like classic, seeing like a state [00:15:00] situation. And then it kind of morphed into this much bigger thing. But Mila, how old, how old were you when that came out? Mila: I must have been in my late when it came out, but I think I only heard about it in my early twenties. I kind of held off of getting one until probably my late twenties. Why? Why? Well, I think I just wasn't at home. 'cause they came to your home to enroll you. Uh oh really? In Delhi, at least at that point when I was live with my parents. And I think I just wasn't home when they came to enroll me. And at that point, for someone for a sort of, you know, relatively, um, rich, urban Indian, you know, adhar wasn't necessary. You know, it was necessary in its early days really for. Indians accessing government welfare or, um, and you know, that was really the use case. And so I'd actually, you know, for a sort of urban Indian who is not accessing government welfare, you didn't really need it until probably the late, the mid to late 2010s. Alix: What was the instigating [00:16:00] factor that led you to get it? Mila: I think I like had to like get a driving license or something. I had to get a driving license and getting my IDs in order was a bit of a mess and I was just like, yeah, why don't I just get this, get my fingerprints in and then it just says that I am who I am. Alix: I'm kind of shook that they went door to door, so like it was like a little caravan of people that would take your fingerprints and like do all your. Like, yeah, Astha: they do that right now as well if you have a baby. So when I have my baby, they came to my house to enroll her, so they take a little footprint for the kids and I didn't have another. Till after I had my kid, because much like Mila, like I needed something. I had been like using my privilege and my multitude of IDs. So many IDs. Yeah. I had so many, I had a driver's license, had a, you know, tax card, a passport and my co-founder at had her Aha. And my husband had another. So I was relying on the ecosystems as well to do very many things. And then I lost my tax ID and you can't renew things. [00:17:00] Without that Alix: because they want you to move over to Adar. Astha: Yeah. And, and then, you know, there were lots of these rumors as to why. And so what was this? This is in 2022. They were like, why in the last 10 years have you not got an adha? Like what is the reason? Because they were like, a billion people have enrolled. Why are you not enrolled? So is there the tipping Alix: point where it goes from. You need this to access government services. Mm-hmm. To everybody's getting it. You need it for more and more things to then you're kind of like a holdout if you haven't gotten it. And then they're like, why Astha: exactly. I use it actually now for very many things, because you can't do very many things without it. So for. Authentication of taxes, both my company taxes as well as personal taxes. We have to do an OTP based authentication. OTP is one kind password. You need it federal like driver's license, any kind of government banking. Engagement you'd need it for. Usha: The very [00:18:00] interesting thing about this project, which I don't think I expected that when the project began, because usually any business has a plan and there'll be something that you'll see, which kind of its own rationality, whether we think the rationality is. Reasonable or not. There is its own rationale here. You find that at every step they were changing track, they were going somewhere else because something was not working. And the interesting thing is that, uh, creating these kind of databases, which are identifying people, you know, the three words that they use. Which were very significant even to begin with. The first was unique and we thought they were saying unique because of biometrics till we realized that they were talking actually about uniquely identifying every person, every individual through a number. So if the number was what was unique, not biometrics, and that makes sense because they didn't know then if biometrics would work. And there is a very interesting history to this, uh, you know, to what has happened since and where we [00:19:00] are today, where basically biometrics is deeply faulty. None of it seems to be working. It began with saying, okay, let's do fingerprints. Then they said, oh, that's not enough. Let's add ideas. Then they said, okay, you don't have to give your, uh, the email address and your, uh, phone number. Necessarily it's optional. And then they said you actually can't function without it. So you can't update your biometrics unless you have a phone which is registered with it. You know, the system is so messed up. And somewhere along the way they said, we'll have a virtual id. Nobody knew what on earth it was. And they said, you must have an other ward so that you keep all these numbers. Say nobody knew, knows what that is. Suddenly one government will say, oh, we are going to have other vault. And I'm like. I thought this was said in 2016. What are you saying? Are you telling us the 2 20 25 that you might have it? So there is, uh, an incredible irrationality with this, along with which there is also this very interesting thing about this project that they do not have a feedback loop. So you're [00:20:00] experimenting. But I can't even call it an experiment because in an experiment you're trying to figure out more and more and see what the fact is. But here it is actually pushing a project, you know, making it mandatory and saying, everybody has to get onto it. We figure out how to make it work or not. That's our business. But you give us all your data. Astha: This is such a wonderfully Indian thing, right? You made something invisible and magical, which is like, oh, your ID is linked to your body. It doesn't need a card. It's just like this science fiction stuff where you can just put your iris and fingerprints and then they realize actually Indians need physical id. Right. So they want that. They want that laminated card with the photo and the number and all of that. And so to Mila's point is that it's been morphing. So the other thing is that it has become a photo id, which it was never supposed to be because you were not validating photos. You are validating fingerprints and IRIS scans. Which are unique. [00:21:00] And so now people have these like paper IDs laminated and they're really long and awkward by the way, which they carry around as like Id proof at airports, for instance. But these are not verifiable photo IDs, even though they're accepted now. As a photo id, the system has sort of been creating space for AHA in ways that was never originally imagined. And as a result of that, it's actually also an interesting study into building something and constantly chipping away at it to meet people where they are. And in that process. It loses its original thing of being just an ID link to your body that you would be able to authenticate where you are. Now you're in this like trap of like photo IDs, OTPs, various other bits and bobs that make it more clunky than it was originally envisioned. Alix: Yeah, it seems both easier to fake now, but also focused on. Whack-a-mole fixing the [00:22:00] latest thing rather than this unifying ID purpose At the beginning. I imagine that the original purpose has not been fulfilled either for state people that work in government and also. You? Mila: Well, I think part of the story is that you see the purpose keep changing or you see new purposes keep getting added on. And this is part of why AAR has been so successful is its ability to morph like this. There's like many books about the history of aar, but in one of them they describe like how the software people, the coders who designed it in the beginning were really insistent on it being this biometric thing and not having a card. And all these people in the government were like, everyone wants a card. The politicians want a card. Like they need to be like here, like we're giving you this new thing. It's also like its success is partly due to. Being able to make it useful to a wide range of people in a sense. Like maybe those promises come before what it can actually do, and then you have to go and make it and do those things afterwards. Alix: You like backfill the [00:23:00] promise. Mila: You backfill the promise, which is, I mean, which is how a lot of technologies developed, right? That's part of it's success is, is exactly in being, being everything to everyone. But Alix: when you say success, I feel like you don't feel great about where it's gone. So when you say success, you mean uptake? Yes, correct. Okay. Um, do you guys wanna get into where you feel like it's not been a great success? People might not immediately understand that the failures of a project like this can lead to such huge consequences for individuals in particular communities. Astha: I think that some of the bigger concerns are that, as we were saying, that the ID gets linked to so many different things and so many different parts of the state. So for if whatever reason, the ID doesn't work. So if it doesn't work for you to authenticate yourself or your driver's license. For instance, you can't get that. It doesn't authenticate for you to get your right to food benefits. And so you don't get that. And in a [00:24:00] country which is as precarious, and let's face it very, very poor, like India, where there are deep dependencies on the state and a lot of faith and trust in the state. As well, right? And you start to mediate the state through a digital mechanism, then there will be consequences, which is that if you're not able to reach the state in the way that you desire, if you're not able to get what you need from the state in terms of your entitlements, your rights, and there are breakdowns there, then there are consequences. And so from the absence of accessing your entitlements of food. To getting left out of lists on pension or not getting driver's license or not being able to open a back account. There are consequences when your relationship, your very intimate relationship with the state is digitally mediated through this biometric ID program. Mila: I mean, I just also wanna add, just score the scale of this. I think there's something like 800 million [00:25:00] people. Registered under the PT s, the public distribution system for food support. Yeah. The margins is still a tremendously large number of people. Exactly. I think we had, we had one study, we had a randomized control trial that showed like something like 2 million people lost access to their food rations at one point during the rollout of AHA in one state for a couple of months. Then they got it back. But you know, I mean, even like that kind of quote unquote teething problem is. Is is huge. Right. But I think we've seen that kinda redresses really difficult to get in those situations. I mean, you have a lot of problems with like getting paid also in these situations. Do you go to your bank branch? Do you go to do your local bank branch? Do you go to like the central bank branch? Do you go like, you know, do you start writing to the main office in Delhi? You know, you have a whole. A lot of back and forth. I think recently there was also a study on like how much time people have to spend to go and get their details updated or fixed at your local adhar office. I mean, the study found that like, oh, like [00:26:00] often the offices were closed like that. The hours are just not clear and people you know will take the day off work, take a long journey in public transport to get there. It doesn't happen. You go back again. If you multiply that by the scale that we're talking about, the effects are huge. Yeah, the anthropologist, NAFI Hassan has called this slow violence where it's not like, it's not like it's clearly violent, but you're kind of in this endless loop of like waiting beseeching, and that's like kind of a kind of totally normalized experience for most Indians at this point. Alix: Yeah, I love the, I mean, not the. Results of slow violence, but the concept of slow violence. 'cause I think in digital systems that is like oftentimes it makes the harm abstract also and hard to explain. Um, because we're used to more punctuated acute violence. Um, that actually, yeah, the kind of harm caused by these things is definitely violence. 'cause someone's making a choice. Usha: It's fascinating that they started with saying, this is basically for the poor, and so that you know the service [00:27:00] delivery is affected. Well by 2025, they're saying, okay, chuck that up. We don't have time with it. Interestingly, also, in 2021, which is three years after the judgment of the Supreme Court, they organized, it's on YouTube so you can see it. They organize a. A program called a 2.0 workshop, and in that you have the first mission director of the UID ai and another person, Mr. Shaman. And Mr. Singh, who was sitting there and talking. And in their session they're saying, who said that this was for the government? This was to create a digital economy. Why is the UIDA acting as if you know, they, they're gatekeepers to the use of this? You should be encouraging it. And this is for a digital economy, so it's played, they're saying it's not for the government. And they say that, you know, we created these three magic things called in mobile being bank accounts, a being. Agar and mobile being the mobile number. If you link [00:28:00] these three up, you can create huge products that is the extent of invasion of privacy, which they are openly talking about in a meeting of all the people who were involved in the development of this project. It's a legitimate question to us. What project was this? Was this a technology project? Was this a governance project? Was this a private industry project? Was it a control project? What is it? Mila: Conceptually, one of the main issues is that it shifts the responsibility for identifying beneficiaries of welfare from the state to the person themself. And so then the sort of costs for making that identification also fall on people. And that's, that's I think a huge conceptual point there. That's really worrying. There were obviously, you know, if it doesn't work, you could work this out at community levels if it doesn't work more locally. On the flip side, there was always the fear because that was the possibility. There was always the suspicion that people who are not deserving were getting welfare, and that [00:29:00] there were huge leakages in the system, which there were. And so maybe there's also something about AHA or systems like Aha, where they're really premised on a fundamental suspicion. And they're kind of trying to separate out like who is suspicious and who is deserving. Alix: I feel like this is the fraud in the us. I don't know if this is a global frame, but fraud, waste and abuse as an organizing principle for state allocation of resources that like you're obsessively looking for fraud and waste and abuse rather than how do we enable human flourishing. Maybe the state takes small risks that there will be some fraud, waste, and abuse, but the remedy is worse than. Fraud wastes to abuse because the remedy creates all kinds of other harms in response. And also it creates this ideology of the government as trying to somehow protect the state rather than protecting people. Astha: I think the leakage question has been part of India's like development for a really long time. Right. Do you wanna describe leakage? Alix: Like what is, what is the, yeah. Astha: Yeah. So [00:30:00] Rajiv Gandhi, who used to be the Prime Minister of India, what was it, 80? 85, 86. And so he has this code, which has been morphed and retold many times, which is that I send a hundred rupees or 10 rupees from Delhi, and by the time it reaches the village, you know it's reduced to 10 pesa. So this particular narrative that there's sort of. You know, people stiffening off government resources at every stage, and if you connect the state with the individual directly and you let go of all the quote unquote middlemen, then the state is going to be able to, you know, service you to the best of its abilities because the state is a benevolent, caring. Institution and all the people in the middle, those guys are the problem, right? And so what AHA does is actually visualizes you as the beneficiary, as the citizen to the state. It visualizes your needs. You are entitled to five kilos of [00:31:00] rice because we know that from your entitlement in the right to food. Or PDS public distribution database. So now I know you are you and you need five kilos. So now there's nothing stopping the state from. Sending that to you, but all the structural things that lie in the middle of all of that, those are obviously not dealt with. So to give you an example, which is more recent, well, in the context of COVID, what ended up happening is that India at some point went from a universal public distribution system to a targeted one. Right? And then now that we have this targeted. Which is about what? 67% of the population is supposed to get benefits during COVID. People moved from, say, Bombay, back to their villages and this, that and the other. We didn't have the ability to move. The public distribution infrastructure. So for some time we had to make all of it universal because actually welfare doesn't move with people, even if people move with their biometric [00:32:00] digital identity attached to them. So the complexities of distribution, of procurement, of the, you know, structural problems around leakages of any kind were not necessarily addressed to the changing of the relationship between the state and the citizen after Adhar. This sense of that these large scale infrastructures can be one. Is that the state? Can somebody called it like the death hug of the government to AHA in 2014? So, you know, how do you also create these large scale systems but leave them slightly outside the reach of the state? Alix: Let's, let's try and like link. The kind of mythology of Adhar and its development with like the next phase of how this kind of escaped containment in India and became this like global, I don't know. Mila: Yeah. So what Alix: happened? Mila: Um, well [00:33:00] first of all, within India, sort of the adhar model of like this platform ecosystem model within the state kind of becomes generalized and applied to. A range of other infrastructures. So you know, you have the unified payments interface, UPI for payments. You have a range of other sort of infrastructures within the state. You have Digi Locker for your broader identification needs that you start getting it rolled out in different sectors. So you've got like, uh, health IDs, you've got student IDs, you've got your epass like toll kind of system, the highway system. Wherever you needed to identify or to track someone, you've got sort of an idea on the Adhar model and this is called um, India Stack. In India. It sort of emerges around 2015, I say emerges. It was proposed around 2015. Um, then you kind of get. This kinda self-conscious promotion of a model or an approach, which I think is, is important [00:34:00] because then it becomes like generalizable, right? It's like, oh, you could apply these principles to any system and remake them. How they're framing this is that like if we build a platform, we can centralize all instances of a particular operation, and so it's much more efficient, like. You don't need to duplicate work across, like every single phone company doesn't need to build their own infrastructure to identify people. You can all just use the same one. Right? And so there's that kind of promise of, of, of uniting all instances of the same operation. And so we have India stack. And then meanwhile, like in the global picture, I think around, you know, I mean we had a sort of shift in global conjuncture where you've got like. I mean, a bunch of stuff happening. You've got the first Trump presidency in the us. You've got the Russian invasion of Ukraine, the 2014 or 20 15 1, and sort of the cutting off of Russia from Swift. That happens then and you suddenly sort of realize that that sort of old, you know, I mean, I think we have, there's like a broader awareness of [00:35:00] infrastructure that happens within the tech world where it suddenly becomes an issue. Then suddenly it becomes kind of interesting or useful to other governments that, that you might wanna build your own infrastructures in-house. So I think that's part of it, like a kind of sovereignty, a sort of sovereignty motivation for building your own digital public infrastructure. You also have, like within the World Bank, you've got the ID for Development program, which is kind of. Pushing the possibilities of digital ID for financial inclusion, for welfare, for all of these kind of developmental purposes. And so I think it's like a confluence of sort of factors that that really lead to DPI being promoted around the world to sort of the emergence of the phrase DPI in the first place. Usha: You can see that over time, for instance, you'll have in 2016 when the law was going to be passed, they said, this is fantastic because now we can be cashless, paperless, and personless basically at that time. And that's the [00:36:00] ambition that. Has been coming, that we don't want to know the people. We want to know what they are doing and what they have. So we just need you to be presence less, but let us know that it is you so that we, you know, we know what you're doing. We'll know what you have, we'll know how you spend. So in 2016, for the credit, we, there was a. A report and Niani wrote the forward to it where he says that people are not, you know, banks are not going to be interested anymore, or lenders are not going to be interested in physical assets that you have. It's about digital footprints. If you don't leave digital footprints, then you know you are not going to get credit. If you don't reveal yourself at every stage, we are not going to trust you and you don't exist. I don't think these were unplanned. I think. As it goes along, they're also getting some unplanned stuff and just experimenting with it. And you know, gates comes back to mind because in another interview he gave two people, they're discussing a range of things and towards the end of his interview they ask, it's a long [00:37:00] interview, they ask him some question to which he answers about India. And he says, uh, in the Gates Foundation, the biggest offices in the US and the next is in India because India is the experimental ground. You know, we do our experiments there. If it works, we take it to other countries. Alix: Does one of you mind replaying what UPI is, what maybe people expected, and then how now it's like Google and Walmart that dominate. Platform. 'cause I think that's an interesting, to me at least, that's an example that really helps bring home how much this is enabling corporate activity rather than public interest maybe. So, does one of you wanna explain that little arc? Astha: So basically up, as I said, sort of was in the works in 2015 when the whole digitalization of payment stuff started. And this was, you know how you have the point of service machines and cards that you tap, charge emergent discount rate that is split between.[00:38:00] The card company, the post company, and the bank that's clearing all of this. So yeah, like Visa, it Alix: gets like 1.5% every transaction. Yeah. Astha: Yeah. So would a, so they were like, this is going to be a seamless thing. So it's an open protocol that is entirely interoperable. So you could have a wallet, you could have a bank account. Only the thing that's not included is cash, but you could have various instruments and then you could seamlessly transact. In 2016, we have demonetization of our bank notes, and so as a result of that, all the cash that people were using to transact, what notes were demonetized? 501,000 were demonetized, which was our highest. Currency. And so as a result of that, people moved behaviors to digital payments and UPI was there ready to be used. So it makes, when you Alix: say demonetized, you mean people, they, they stopped being like you couldn't use them anymore. Astha: Yeah, overnight. Okay. Alix: Yeah. [00:39:00] Um, whoa. Okay. Astha: Yeah, that's another episode. But yeah, so it was a real shock to the system and so that changed behaviors for many people. Even though the arc of UPI payments, you can see that it, you know, picks up. In the months and weeks after demonetization, and then people start to move to cash again. But then the habit really stuck was around COVID, right? Like in 2020 you had to, because you're just transacting digitally sitting at home. There's the, you know. Proximity issues, people don't wanna touch money, whatever. All of those things. So digital payments, and then UPI becomes your ability to just scan a QR code or take somebody's number. So you don't even need to give your back account number. You can just have a payment ID that you can use to transact. So it's really swift. It's really easy. I mean, yeah, it takes no time. It's real time. The recall is immediate and that's the. UPI story. And it's a very interesting one because [00:40:00] again, like the Indians, they thought that you could just make these magic payments from, you know, you'll scan a QR code, the money would go from your account, and then people didn't really fully understand how money moves. I mean, even, I don't understand how it happens. So, so, you know, they, they had to. Almost evolve a little voice box with one of India's most familiar voices, which is an actor called AM Bja, and he recalls to say, you have now transferred X Rupees has been received in terms of what happens. So. UPI is this open protocol and the objective was everybody is going to be able to build payment applications on top of it. The Reserve Bank of India had a beam app, which is uh, was the sort of public option. There's an Indian one with phone pay. It was like an Indian startup. And then the third one was Google. There was also of pay TM, so there were four, and Google Pay was the first Google product. Developed outside the us So it [00:41:00] was homegrown on India, on the UPI protocol, and then cut to now, uh, PTM still remains a payment option, but phone pay was acquired by the Indian version of Amazon Flip card. And then flip card was acquired by Walmart. So you know, phonepay is now owned by Walmart. And then Google obviously sort of entrenched itself through this. And so now you have about. 85 to 90% of the payments in the India ecosystem that are owned by an application that is owned by Walmart and an application that is owned by Google. Um, and the public option of p just sort of died somewhere along the way. Mila: It's still there, just nobody uses it, but I use it. Because I don't want to see ads in my payment system. Uh, and so I, I use it, but it doesn't work as well as the others Alix: would. They get you both ways? They take the transaction cut and they serve ads. Mila: There's no [00:42:00] transaction cut. This is the sticking point. So there's no actual way for these companies to make money without cross-subsidizing and somehow, like there's a sort of conflict playing out there where you've got the reserve bank, which is trying to promote financial inclusion in state. They, they want payments to be free. And then you've got the sort of payment processors who wanna be able to take a cut. And so there's like a bit of conflict there as well. But, but what that means is that, you know, the reason that the Google and Walmart. Players could become the biggest is because they could afford to cross-subsidize it, because Google's in it for the data business. Right. They're not in it for the payments business, and so that, that's also part of the story. Alix: Yeah. Having access to digital payment data of hundreds of millions of Indian consumers is I imagine very valuable. Usha: There's been a growing idea for Santa now that the private sector can deliver. The public sector can't, but we've had had about 25 years of experience of that, and we know that the private sector won't, that's not their interest and that's not why they've been set up. I mean, there is enough work that's been done on the idea of the corporation for us to [00:43:00] know that that's altruism is not their game. Which is why when you know, when you have, uh, one, this is the other thing that worries us too. Here too. India too, like in the US the big philanthropists are people from the world of technology and so they control the discourse and the idea that the people who might be posing the MA maximum threat to us are also the people who are going to fund us and fund all thinking and fund all work. Is a very scary prospect. I think we really need to have many more discussions on that. We had a so sociologist writing recently in the papers who says that, uh, the way technology is getting deployed around us when it comes to, for instance, food, which will be delivered only if your biometric will work every month. We have work workplaces where they want you to be photographed and. You know, for it to be sent and when you go to work and when you leave work because they don't trust you. So he says, you know, that we are living every day. They say, if you don't [00:44:00] link something up with something, then you're won't be punished. So he said the idea of living with fear every day, you fear what may happen now if that is what technology produces. Just think about what it means when they, when you say efficiency. When you say the good AI for public good, and you're actually every person is feeling like, oh God, what will happen to me today? Will my bank account be shut? You know, will I be able to travel from one place to another? You don't know. If something doesn't go, will I get my food today? You know, will I get the work that is? Something that the state is supposed to provide to me, and especially because you see people around you who are struggling with it. It's not a, this is not hypothetical, this is, you know, observation. You see people around you struggling with it and it puts fear into you. And he says, is that really a society you create? Astha: We were so intent in the democratization of payments, which is that we want everybody on the platform making [00:45:00] transactions that we didn't. Think through. Some of these things, which is that, you know, the state will want to track transactions of a certain kind at some point, right? What does that mean? Or that this is not going to be sustainable for banks even when you have a trillion transactions a day, right? That's the number. So this is still not gonna be sustainable, so how are we gonna think about this? And I think that's what the other interesting part of all of this is that the business model question in the context of. Digital public infrastructure is often unanswered. And there are many research papers around the financialization and how that's going to work. Till now, it's largely been the state or philanthropies or you know, like there's no business model, which is also quite interesting. Alix: That's super interesting. Yeah, and I mean, you introduce something cheaply and then you have to, you get market share and then you have to find a way to monetize it. And I feel like. How are they gonna monetize that transaction data? I'm very, I'm very, [00:46:00] or how are they already, because I'm sure they. Already have found ways, um, is, uh, yeah, Mila: I I might have a slightly different take than us, Todd. This, I think that's what we've seen in India because the state has been so intent on scale at all costs and because their intent on making it, uh, public private play. But I don't think that's necessarily the case for, like, that the DPI systems necessarily have to do that elsewhere. And so like in Brazil. Neither gov B nor picks really have that issue because Pix just charges a small fee. Um, you know, in India you couldn't charge a small fee also because, you know, Indians wouldn't pay a small fee probably. And because, I mean, I, I guess like the starting point was quite different, which is to say that more Indians were unbanked, they're a lot poor, mean a lot more sensitive to fees. But I'm not sure that this part of the DPI logic needs to necessarily carry across to other countries, or if, if such simpler systems are implemented elsewhere. But in India it's definitely the case, Astha: but also just to add. And [00:47:00] I think is that because India went first, and so we know there are these bumps around the road, so it would be prudent not to repeat some of this. So I agree. I don't think that, and also I think that Brazil or Singapore or Estonia were just doing their own things. Right. And building their own systems in the ways that it made sense in their own context. Um, and the DPI tag is, is something that has been attached to systems that they were building. In their own cortex belatedly. So there's also that, which is that we have a certain version of how infrastructure at scale has been developed and what those sort of model standards and ecosystems look like. And then other people. Have built it in the way that makes sense to them, and then now we're trying to stitch it all together as a broader narrative arc. Alix: Next up, um, we are going to travel over to Brazil and look at two other gigantic projects [00:48:00] that have gone. Kind of sideways and interesting and informative, uh, ways. The first is Gov br, which is essentially Brazil's adhar, but it's obviously different in really interesting ways and we're gonna hear about how it can go horribly wrong. And we're also gonna hear about Picks, which is a new-ish state enterprise in Brazil. State ish also, um, that is for making payments. It is essentially digital payment infrastructure, and it is a doozy of a story of how it came about and how it's different than basically any other country's payment processing. System, and you might think that this is gonna be boring because it's about banking or something, but it's actually like full of geopolitics, full of political intrigue, full of some really interesting tactics from the Brazilian government to prevent global North domination. Luan was kind of there. For a lot of it as part of a policy advocacy organization that was trying to make sure that [00:49:00] ultimately, uh, PS was functioning in the public interest. So next week we're gonna dig into gov, br and PS with Rafa Zanatta and Lua Cruz, who are like two of the foremost experts on these topics in Brazil and have some amazing stories to tell. So I hope you, uh, stay tuned next week. Um, and, uh. Hope you enjoyed this episode. Thanks to Georgia Iacovou and Sarah Myles for editing and producing, and thanks to our guests for being a part of not just this single conversation or this single episode, but helping us kinda think through how to put all these episodes together to tell a much.