 This is the final day of excellent week of programming around non-personal data that has geek has been convening. And I'm just here very quickly to introduce the first speaker this evening for the next sort of hour and 20 minutes. We have first Srikanth talking to us about this particular concept around a debt and NPD that he's going to be talking more about. And then for about 15, 20 minutes following which within the 20 minutes slot by 6.20 I'm going to provide some reactions to his presentation and invite comments. We'll have a short break and at 6.30 then we'll begin a moderated discussion around privacy concerns in non-personal data, the V2 of the report with an excellent lineup of panelists who will introduce themselves. So given we're quite short and snappy with time, I'm going to go off video and hand over to Srikanth to take us through his presentation. Thanks Srikanth. Yeah, thanks Malamika. Hello everyone. I'm Srikanth, part of Cashless Consumer and Consumer Collective on Digital Payments. What we'll be looking today is around the topic of debt, NPD and ownership to play infrastructure. Now I'll unbundle each of these as we go through the presentation. A quick overview of what we're applying to cover it for the next 20 minutes. These are essentially political economy observations looking at payment systems and other data flows in them. Specifically, we'll be looking at three payment systems, Bharat Bill Payment System, Fastag and National Common Mobility Card. We see what the non-personal data aspects of these payment systems. And the emphasis here is not on privacy. So that's a separate debate. We've kind of had through all the privacy conversations and there are probably enough literature by now on privacy issues around each of these. So I'll not touch much upon the privacy aspects of these, but we'll focus on the NPD aspects, so which is the non-personal data. So on that, much is started on the economic value of non-personal data. And even if you've seen the reports, there is an economic growth aspect of non-personal data that keeps coming around. We'll kind of see that slightly closer with these real-world examples. So for that, what we try to do is we'll see, this whole notion of NPD has come around in, say, 2020 when this committee was formed and it produced two versions of the report which we are discussing through the week. But we'll kind of see how this essentially non-personal data was still kind of managed on these three payment systems. And for which we kind of see what's also common is all these three payment systems are essentially part of the indicted infrastructure sector. So you might have heard in the news about the non-performing assets of banks and so on, which basically the loans that are not coming up. And these are primarily on the infrastructure sector. And when we say infrastructure sector, one is electricity. That's where PVPS will come in and we'll go through what it is. Next is FASTAC, which is where all your highways and tollways come in. And then the third is National Common Mobility Card, which is where all the infrastructure around your metros, metro rails in urban context comes in. And we see how all these three have been having huge debt issues or their financing of infrastructure and what these systems and the NPD around them try to achieve and what where we are today, we kind of see a brief overview. And then we'll kind of use this knowledge as a background and see what NPD is. Because mostly you've kind of seen NPD from the prism of let's say startups or as a community is generating data, wanting access to data. But we'll kind of see what NPD is through the lens of the state or what NPD entails to the state and what we should be worrying about. And as always, standard disclaimer, views are personal and of course, political as well. You may agree or disagree with me. So we'll get started. So first is power sector. What is the role of BVPS and NPD? So a brief history of power sector reforms in India, I think ever since we've kind of opened up our economy, there is a significant push in privatizing the power sector which is traditionally held by the state, both when all across the various forms of it, so be it generation, transmission or distribution. And there were initial set of reforms which kind of uncoupled each of these functions and said we've kind of largely allowed range of private sector players in the generation aspect. And in some cases, transmission and distribution as well. Although distribution is still held largely by the state, the individual state governments own electricity company and that has also led to some kind of a financial crisis in them, which is why you may have heard these names, Uday scheme that was kind of put in place somewhere in the NDA one, which is to kind of bring these distribution companies to a better financial health, relieving them of their excessive debt burden. And one aspect of it is, and which we are now seeing as part of the Atman-Balbarath is the Discount Privatization. And we'll kind of closely see what's the role of MPD in all these. And so one problem is essentially, so we've kind of seen one part of this puzzle is that Discounts are unable to collect electricity bills. And so collections have been a problem which is kind of leading them to a place where they are financially unhealthy. So one, there has been multiple committees and task forces between thinking around solving these problems. So the initial, the early reference to that is the 2002 power sector IT task force, which kind of suggested some, it's a computerization of the consumer information system as well as bill collection system. So let's say the earlier versions of digitization of these to bring in efficiencies and see improve them. And then slightly later in the day, RBI came up with this concept of a bill payment system. And in 2013, it kind of floated this idea again. And what it essentially meant through that regulation was to create an infrastructure where RBI mandated all utility bill payments to be centrally routed through this system called Bharat Bill Payment System. So this is basically, so if you have kind of a brief history on payment systems on utilities is essentially you would kind of be going to the counter and paying up in cash or probably website of this utility company. That kind of changed over last three, four years when since Bharat Bill Payment System was gone live, you might have seen your Discom getting listed in your favorite app, payment app, and then you could just pay or get the staples. That's possible because of BVPS and BVPS is mandated regularly by RBI. What BVPS also does is basically creates a centralized data source of all the receivable data of all the discounts. So essentially, we can practically say that every electricity bill in India is being centralized in this BVPS data store. And this, of course, does have its own personal data implications. I'm not gonna go into that. We're only gonna see the non-personal data implications of that. So on the non-personal data front, how can this data be leveraged? So one is as part of this privatization they are planning to kind of boost this data and run some data analytics and then see how we can kind of segregate these companies into more efficient units and then privatize them. So let's say today your one big state has one single distribution company. Now how can each city have its own distribution companies and how can they be privatized individually so that their value can be maximized and so on? So this is one place where this data, the receivables data of discounts is being used. So this will probably go into valuation of these, fitting of these discounts and so on. Let's say when the other, why this focus on, say, sizing discount data. The other reasoning is also that there's a huge transmission distribution cost in kind of distributing power. So if we have better knowledge as to what the consumption patterns are down to the, let's say street transformer, then they can optimize, the distribution companies can optimize their networks and then use and deploy their infrastructure effectively to kind of save load and they can analyze consumption patterns across various regions or areas in the city and then they can kind of manage their network efficiently. So that's the other role of non-personal data in say power. And what we've done through VVPS is essentially, we've kind of centralized this data infrastructure at a national scale. So VVPS as with any other payment system is being run by the National Payments Corporation of India, which you might have known that it's a bank owned segment infrastructure company that's collectively owned by a set of banks. There's also been some recent discussions around forking off a subsidiary for VVPS. Now, this all discussion is of importance because if you see the value that MPCI, which itself is a nonprofit has and is only data and this data is not just the personal data of, say you and me using UPI or the other payment system, but it is also this non-personal data forms a crucial role. And that is forking that off to a different entity. One kind of needs to see who benefits out of it. We leave the who benefits out of this just for your thinking. And I want to kind of bring in the NPD paper context here. So the NPD paper context here is essentially a data trustee. So I mean, if you read through the NPD report, you notice the data trustee being referred and here data trustee kind of aggregates the NPD of the entire sector. So this subsidiary for the VVPS will kind of own this NPD of all the power distribution companies. Now, we'll see what's the issue with that towards the end. Now, this is power sector. And then we have a similar set of issues on road infrastructure and FASTAG, why we are having FASTAG. So again, brief history of highway development in India that being like multiple PPP modes of building highways ever since say the Golden Quarter lateral days and so on. But even that sector again, faced significant financial health issues, which means that say the contractors who are putting the highways were not getting compensated or the state was losing out a lot more money, tools were not collected properly. So FASTAG was kind of again conceptualized as an electronic tolling system, which will kind of maximize the revenues of. So just like how in the power sector, we had a collection problem. So even in the highway sector, the tolling was again a problem. And FASTAG was kind of pushed as a technological solution. It was conceptualized in 2010, there was a committee report and it went through multiple iterations and kind of fully went live in 2020 and it's now managed in 2021. Again, FASTAG has its own say the privacy issues, we're not gonna go into that, but let's quickly see how the NPD role of FASTAG data is. Now FASTAG kind of collects, let's say the traffic volume of each highway systemically and with which, let's say the traffic volumes across individual highway segments can be measured in very absolute terms because especially if you make FASTAG as a mandatory mode of payment for tolling, you're basically having the entire tolling infrastructure data in a single place. And if you strip off the personal data part, you'll be having a rich NPD data set, which is then further used to decide policy and the policy here in the road infrastructure has been like two financing options. So one is called a model called toll operate transfer where publicly funded highways, which is basically highways built with the government money or in some cases, if let's say a highway was built by a private player, but the concessionary agreement had completed over 20 years. Now that kind of, that highway comes in with the hands of state and then state decides to monetize that asset, the highway as an asset. And while monetizing that asset, the state is using these FASTAG data to kind of value what the highway is and what its potential is. And then it puts a base price upon which multiple people bid over that highways for tolling rights. And use that and then they participate in an auction and they get the tolling rights and then they continue to toll and state gets a one-time monetized value of that asset. So state no longer owns that highway, now the highway is kind of sold off. So that's where an NPD is crucial there in the context of pricing the bid. So how do you value a highway for a 20 year period? So the answer to that is state, I mean, one way is to use the FASTAG data and then do some projections. The other is, again, invits, which is slightly a different variant of this wherein an individual kind of issue like a mutual fund and people can buy invits of the market and the entire highways will be owned by invits. Again, a set of highways will be transitioned into this trust, which will issue invits. So again, the role of NPD here is largely around pricing of these taught and invit bids. And again, housing this data, currently it's housed with NPCI, which is implementing the FASTAG ATC project along with an infrastructure consortium, IHMCL. This again, bringing back to the NPD debate, these are the data trustees of this data set. Again, I leave the crew benefits to you and then come to urban mobility. So urban mobility, again, we have a similar set of financing issues over funding metros for each of these cities. And metros typically take a very long time period to kind of build the entire infrastructure and then ticketing forms will be a very low part of repayment to that kind of debt. So NCMC has been like ideated multiple times, started in 2010, concept slice in 2012, launched multiple times in very different names. Previously there was an entity called UTIITSL, which was owned by Ministry of Finance, which ran some common mobility cards. Currently there is again another pilot by NPCI to do a common mobility card. So every now and then you will see a common mobility card. So the role of NPD here is involved on two fronts. So one is the standard things that we've seen on valuing what the economic value of it in terms of transactions. So how much are people transiting or how much are people consuming electricity or using the highway? Like how much are people using the transit infrastructure is one part, but this transit infrastructure data can also lead to, let's say, planning input into urban development and planning transit networks themselves. So where do you build the next metro line and so on and so forth? So this again, there was a huge value on this non-personal data and then this NCMC kind of tries to centralize this data into again with NPCI and I would say taking away the autonomy from the cities. I'll again leave the who benefits to you. But so this has been the case with India and how India has kind of navigated through this for the last decade, but we'll kind of see some examples globally. So there is Saudi Arabian bill payment system which is called SADAD is being run by the Saudi Arabian banking regulator themselves and it's a fully centralized data store, an NPD, much like the database itself is with the regulator of the state. And I can loosely call it to, let's say, a Marxian model where the state owns all this data. And then there is oyster cards in London. So they have a transport for London authority, this corporatized entity looking at all aspects of transportation in London. They again issue oyster cards which is for the transit payments in London, which again creates a centralized data store which also generates NPD around transit patterns in London and so on. It's with this entity, TFL, but it's also TFL again rolls up to the London city council and so on. And it's sort of, let's say NPD is accessible not just to the state, but lastly to the community and the market as well. And I would say broadly this I'm equated with the Keynesian model. And then we'll take an example of tolling and toll interoperability in the US. So US also has a freeway system and multiple tolling, electronic tolling projects much before we've had. And there again, there was a concept conversation recently a few years back around why do we need to have a tolling, e-tolling infrastructure uniquely separately for each state? Why can't we have an interoperable tolling system? And there, the way they've solved it is not by having a single entity, but then used a federated interoperability standard with which you can still move across multiple toll providers and still have different set of, I mean, you don't need to actually have multiple tags or transponders and can still transact in an interoperable way. And here the NPD is kind of collected in a federated manner as in whichever tolling company owns that particular highway kind of has its toll data. It's not shared with the state. So it's kind of NPD is with the market. And I kind of say loosely say this is an Adam Smith model of housing this NPD data. And India we've just seen through PVPS, NCMC and NETC which is fast type. We've kind of have a slightly mixture of all these put together in one form or the other. Now, we'll kind of broadly see which are like the NPD generator systems or high value data sets systems currently available to kind of before we see what NPD, what role NPD plays and I've kind of split this by ownership. So let's say the first generation systems or the legacy systems, we've had say Waffen, which is the vehicle data. You might have heard this in the context of privacy leaks or a Waffen data sharing framework and so on. And then we have GSTN, which houses taxation data, PFMS has public class data. ENAM is an agricultural market, but it's basically run by the state. And then IMPDS is basically the ration shop network. So you, which gain houses, both food consumption data as well as potentially migratory data as well. And then say we have e-courts, which houses all the litigation data, NCRD houses crime data. So this is all basically state owned or state built data infrastructures, which have a very high value data sets on NPD. Now there is again, moving on like there's a second category, which is regulator owned or built data sets. Examples here would be like the insolvency information utility built by the IBBI or approved by the IBBI. Rohini is a database of all hospitals in the country run by the insurance information bureau, which is again under the IRDAI. FOSCOS is basically all the, it's a system basically that has a database of all the food licensing related data. So all the FSSAI, which is the Food Safety and Standards Regulator in India has a system called FOSCOS, which licenses all the packed foods and so on. And then let's say the trial, which monitors the internet and telecom infrastructure has an app called MySpeed, which is to monitor all the internet speeds and then it kind of houses that infrastructure in itself. So these are some of the HVD infrastructures built and owned. I'm there for you Shrikant. We are coming up to time. So you have a couple of moments. And then we have like the privately owned built infrastructure where we saw these three, but there are some more as well. What do I mean by asset light sector control infrastructure? So we've kind of seen through 2010 through 2020 state built several data platforms, as we've seen in the examples, through a private entity to centralize sectoral data. And this centralization of sectoral data was used with the strategic intent in raising capital to offset debt. And why do I call NPD as an asset light? Because here the government which is not, will not even be building that kind of a data infrastructure to generate that data. And it's also partially because government wants, doesn't want to kind of regulate the digital economy same way, but still want to have some amount of power or control of the economy. So what it means is basically anybody, what it brings through this employee framework is a data trustee which will kind of operate the data infrastructure to collect managed data. And then that is the exit of the government for whatever reasons it wants. Now, what does NPD mean in this context? So NPD will also be an infrastructure insolvency resolution tool. So as you've seen that these sectors were kind of indebted by a huge margin. So the state kind of decided to intervene, put in these data platforms, extract data, generate NPD out of through systems. And then the state is currently using that to kind of get out of the debt mess that it has created. Now, what did to potentially mean this with NPD is state will not even be building those systems, but state still has access to such kind of granular data, which can then be used to kind of decide policy for each of these sectors. And this has significant impact for the existing and new investments across sectors. So we've kind of seen only the infrastructure sector and how payment systems were kind of used to generate NPD for the infrastructure sector. What this also means is approaches to data minimization will be structurally discouraged. So the state says that you've got to generate data. And if you're operating in a sector and the state comes and says, this is the data system and you have to share your data, you might even as an industry, you might not want to actually even collect some data, but you would be forced to kind of collect that data and then give that NPD to the data trustee because now the NPD framework demands such kind of data to be shared to the data trustee. So data minimization options will be kind of structurally discouraged. And then the other issue is around governance of data trustee, some of the framework mentions of data trustees, now what is the governance norms of these data trustees, transparency, accountability, provisions of these and how it impacts individuals, businesses is something to worry about. And that is all I have. Thank you. Great, thanks a lot for that. Actually there was a very, very quick and a nice overview of all the thoughts that you had. What I'm gonna do very quickly is just a way to check if we have any questions. I don't think we do, which is great. We are right on time. I'll just share a couple of observations maybe and then we can go straight into the next panel if that's fine for everyone in order to stay on time. So I mean, just in terms of responding to the presentation Shikant, I thought it was very interesting. Some interesting dots connected for me that I hadn't previously thought about in that light. I think everybody's talking about this foundational question of whether at all the state can do this and monetize data in this way or set up a regulatory framework. But I think two or three specific things I want to call out is one is this presentation. It started doing this. And I think if you're planning to write it up as a paper, it might be really helpful to just call out the fact that you're thinking about NPD in the context of asset monetization. So obviously with Deepum and all of that, there's this huge conversation or an asset monetization that's happening. And I don't think they really thought about it seriously in terms of how do you unlock intangible asset value from data, whether it's the NPD itself or how do you use NPD for business intelligence or when the government is figuring out how to auction off his assets or whatever Deepum does. So I think that's something that it's in there and I would love to see more about that maybe in a future presentation. I think the other question that it raised for me is I would like to understand whether you're thinking about NPD in itself being valued and in some sense that itself is the public asset or are you just thinking of NPD being used in these kinds of valuation exercises? And I think why that distinction is different is important is because in the first where NPD itself becomes valuable, then are we starting to think about it as public infrastructure of some sort that everybody is getting open access to and then a whole new range of regulatory and other issues come up. But if we're only thinking of it as business intelligence, then of course you're essentially thinking how can the government use this better or how can private players like you had in your table use it better. And quickly, I guess the last thing I will say is just I think there was a little bit around financing in there and a little bit around the regulatory, what's the intended impact of this framework? Again, I would love to have that kind of drawn out. I think you wanted to say something and then we can go straight into the panel discussion. So I don't consider the NPD as an asset in itself. What I meant as an asset-like infrastructure is basically for doing this, the state had to invest significant capital, political and economic capital to build out these data infrastructures. So we've talked to these three and it has taken a decade long. With NPD framework, this thing is kind of going is it's lashing on to the industry and saying that we are not going to even build this out. We want you to give us data that we want and then we will use that in policy making. So this is like going, if the asset-like framework is basically, you don't even have any asset or data infrastructure and the only thing that you're putting in place is a NPD framework, which will kind of latch on to the existing data sets by various industry, community, whatever, by mandating this community data sharing. And this is just one view as to how the state could potentially view. I'm not saying this is how the state intends to, this is what the policy intends to, but this is possibly one consequence of the NPD framework.