Decoding Universal Health Coverage with Dr. Nachiket Mor: Part 2

Introduction: You are listening to Decoding Impact, a podcast by Sattva Knowledge Institute hosted by Rathish Balakrishnan.

Welcome to Season Two of Decoding Impact. Every fortnight we will engage leading thinkers and practitioners to understand what it takes to solve systemic problems at scale. For all the curious changemakers committed to understanding the trade-offs and incentives to make this world a better place, this one’s for you.

***

Rathish Balakrishnan (RB): [00:00:48] In the 75 plus years since India’s independence, a nation has evolved, but has our imagination for this country evolved accordingly? Especially when it comes to health care, where the health care burden on this country is far more significant than where it was 75 plus years ago. In Part One, Dr. Nachiket Mor and I, after setting the context for universal health coverage in India, today, discussed what are the type of mental models needed to fundamentally build trust in the system, the necessary infrastructure and human resources required to enable accessible and affordable health care. The importance of primary care and shifting the governance approach from being top-down to addressing individual districts as units. If you have not listened to part one of this conversation, you can find it on our Sattva Knowledge Institute website or wherever you find your podcasts from. In part two of this fascinating discussion, we’re going to look at the potential of healthcare solutions that are emerging in the Indian context, the role of the private sector, the all-important question around financing in healthcare and more. So welcome back to Decoding Impact and we hope you enjoy Part Two of my discussion with Dr. Nachiket Mor on the topic of achieving universal health coverage in India.

Rathish Balakrishnan (RB): In the part two of this conversation, I wanted to spend more time on things that are emerging now or questions that we still don’t have clarity on. Maybe there is a greater haze about this and maybe there is a need to wait and watch. I wanted to start with digital public goods. This is the G20 season. There’s a lot of celebration globally – and I’m seeing that first-hand – around this idea that digital public goods can serve as a way to usher greater growth, faster growth and so on. Everyone from Bill Gates to the country leaders globally have recognised this. What are your thoughts on how does it fit into the plans on accelerating universal health coverage?

Nachiket Mor (NM): [00:02:54] So public goods – you take away the digital notion for a minute have always been seen as important. If you look at the history of welfare economics, the original thought was first the welfare theorem, which is price-based. Markets will figure things out. What is the role of government welfare and maximising government? Do nothing. That’s the role of the government. Quickly people realised, “Well, that’s a great idea, but the equilibrium that is emerging from markets may not always be what we call socially optimal. It may be Pareto optimal.” Well, people said “Simple idea, why don’t we do this? Change initial endowments, tax the rich, hand over the money to the poor, then let the first welfare theorem operate again.” Very soon, people realised that there is a missing piece here. Let’s take laws. Now the markets need laws. But the problem with laws is that they’re not limited in supply. You cannot deny anybody access, which means they won’t have a price, which means market can’t manufacture them. If markets can’t manufacture them, who will manufacture them? But without that, the markets can’t work. This way, the notion of public good started to emerge to say, “Oh, that’s another role for the government.” One role for the government is do nothing. Second role for them is tax the rich, hand over money to the poor. The third role is to build public goods, core role of public goods – help markets, because without them, markets can’t function.

NM: [00:04:12] The underlying belief was markets will do everything else. That is the mechanism. You know, if you take the language of game theory, there is a Nash that will get built without a public good. With a public good, like good regulation, the Nash will shift. You know, you’ll move from less cooperative game outcomes to more cooperative game outcomes. Clearly, there is a role for public goods, digital or not. Now the question is -is this new phraseology of digital public goods, the core of Public Goods still needs it and in some places technology can play a big role. My fear a little bit is that this phrase is being applied somewhat more broadly or more loosely than this conversation about shifting the Nash, or is there a real role here, for example, I was in another discussion yesterday in which somebody was talking about digital public goods in health care. One question I said, “What I don’t see in your discussion is where is the market failure? Because if you don’t tell me where the market failure is, then I don’t know what exactly is your public good trying to solve? If the markets are not failing. Do I need you in there?” Now somebody will say, “Oh my God, it’s free.” What you’re saying is, it’s free because somebody else paid for it. Now think of it from a national equilibrium argument. Some philanthropist volunteer decided to pay taxes to themselves, not to the government, because this allows them now to take that pool of money and to do what they think is the right thing to do with that money.

NM: [00:05:50] Now, sometimes that could be a good thing, sometimes it may not be. Sometimes it may hurt a market, when markets actually are going to solve it. It doesn’t meet the test of a public good. It’s a private good that somebody decided to pay for. If the public sector offers C-sections, it’s a public good. It’s not. It’s a private good. It has a price. Markets produce it. The government decided that they are going to bear it through taxation. You know, and as some academics have pointed out, taxation is one of the most expensive resources. It costs us almost 10 to 15 paise to collect every rupee of tax. We’d better be careful what we do with that expensive money. So that I think is a piece that needs more clarity, where we’re not saying “Therefore they are not useful.” They are important, but we have to ask the questions – what is the Nash being shifted? Which is the market failure that we are trying to address and how are we making sure that this is indeed the market failure we want to focus on and not being driven by some voluntary tax payers who are deciding to now assume the role of the government and say, “I will pursue this idea versus that idea.”

RB: [00:07:02] I want to speak about this specifically in the context of ABDM and share with you my limited understanding of how I see the value unlocking, the reason why thinking of it as something that the government can anchor makes sense and it will be good to get your thoughts. One is there is always the challenge of data lock, in that many countries have now seen where private providers get access to a large amount of data and the data itself is incredibly valuable but does not always serve the purpose of delivering better public health. The point you made in our first part of the podcast about where the opioid crisis is actually emerging. Where do we see larger number of HPV cases rising? Where do we see malaria rising, for example, while there are satellite imagery et cetera, if there is a way, there is a common backbone where the nation can look at health data across the board, there is value in it. But if the government were to build a data system and that’s not going to work, and that’s been one of the touted value propositions of ABDM, which is to say, “Hey, it will allow us to create a common data backbone where the data will help make better public health choices.” Because we finally have population scale data which is not survey based, but it’s more real time and so on. The second argument that has often been made is that the technology infrastructure removes the transaction friction on things like insurance, claims transfer, processing, etcetera, which today is one of the reasons why, despite the spend allocation that we have, there is underutilisation, because private markets don’t have the incentives to be part of these schemes because the payouts take so much time and so on and so forth. Saying the technology backbone that identifies every health provider can enable this friction to be reduced, which then enables them to deliver better value. That’s been the second one. The third argument that in general we’ve made around digital public goods is accelerated innovation. If, you know, like for example, in agriculture, they say that “70% of money spent by venture capitalists is on enumerating farmers.” Same is true for health care as well, and so if you sort of enumerate every health facility and have a channel to reach, so many doctors are providing a starting point for private innovation to build on top, to enable discovery ratings, quality control, all of that stuff which private innovation might find a business model to actually solve for. These have been the three large value propositions that ABDM makes the case for and I’m not essentially making our argument if it is a good thing or a bad thing, but the point you made, if this is the value unlock, is the articulation of a digital public good, the right articulation for what we are trying to do.

NM: [00:09:24] I mean, all of what you mentioned is valuable. I’m not going to dispute that issue. The question is, are these the market failures? It is good to have insurance claim process come smooth, but it’s better to have insurance. You have only 6% of the population insured. How are we going to solve that problem? The government already has an insurance scheme. It’s called free health care. What about that? Why have you chosen a tiny portion of the overall financial ecosystem in insurance and said, “I’m going to fix that”? That’s not the priority. You have a bigger problem. For example, if I look at a state like Kerala, one of the issues that we are struggling with there is that the Kerala underlying government health system is not automated. So forget automating all providers. What about the government itself? Now you could say, “Well, this is a way to do that.” Why not? But that’s not the articulation here. The articulation here is a broader articulation, which in some ways could be good if were to get the sense that it’s driven by solving the narrower problem, but it seems disconnected to it. So that’s where a little bit of my concern about this rises. Completely agree with you that ability to capture population-level data – if look at, for example, the British surveillance system, the best surveillance system they have is the general physicians, a chosen group, their data being aggregated automatically at a national level to understand real-time trends on fever on this, that and clearly there is value to it.

NM: [00:11:03] Now, there are 2 or 3 concerns there. One, the country is very large. I don’t need everybody’s data to draw meaningful conclusions. I need a sample. I don’t need to know every single individual’s every single record to understand that malaria is growing. In fact, the worry will be that now you have not built what we call sentinel sites, which are there for the explicit purpose of collecting in-depth information. What you’re relying on is fairly shallow engagements with patients at curative points. To aggregate data, you might miss the real story until it’s too late. I’m not saying therefore this is not a good idea, but I would say it doesn’t take the substitute, the core public health idea of going and sampling sewerage water to check for the polio virus. You’re not going to solve that through automation. Some human being will have to go there and take that sample. Some people will have to go to the deep forest and look at bat droppings to see how many bats are there. Some people will have to go to homes of people to count the number of dead mosquitoes in order to understand what is the flight range of the anopheles mosquito from the stagnant water farm site that you had? So these are good compliments, but don’t want anybody to assume that we have this so we don’t need that. You’re going to need both. There’s another big concern about privacy. Even though I didn’t get your name, I got your DNA, I got your disease profile, got everything else.

NM: [00:12:37] Deanonymization is trivial. I need you to touch the system in any place, in any capacity, and I can quickly say this belongs to this individual. It may have nothing to do with health. Health data didn’t have the name, but had everything else. The question is, why do I need all this information? Actually, it turns out much of this information is not needed. By creating a health system which is solving what I would call not a market failure, you could open up new risks. So that would be my concern. I’m definitely of the view that building this kind of structure, so that repeated similar investments, like in data gathering efforts should not happen is an important goal, but would have started first by trying to understand which don’t see the articulation of what is the market failure in healthcare. Efficiency is a concern. But the real problem – why are we getting misspent money in health systems in states like Kerala or Delhi? Overuse of hospitals, how is technology going to solve that? I need better primary care. I already have primary care. I need people to go out. I need all of the stuff we discussed earlier. Technology is not going to solve that. If I don’t solve that, I don’t solve the inefficiency problem. Even if I put the latest, greatest glitzy machine, not all friction is bad. For example, without telemedicine, I was forced to go to my neighbourhood physician. Now, instead of having the patience to improve the quality of my local service, what you said is I’m going to make you able to access anybody in the planet.

NM: [00:14:21] Great idea for secondary care. Great idea for tertiary care. Terrible idea for primary care. Because now who’s going to worry about did I take my medicine or not? Who’s going to worry about my social determinants? Who’s going to worry about my tribal ritual versus my health? Nobody. This two-minute conversation – because the primary care image behind that conversation is primary care equal to mini hospital. As we discussed earlier, it’s a very narrow part of primary care. So I would say, therefore, this architecture is a great idea. I would spend a little bit of time thinking about which market failure is it solving and how do I go about – not necessarily changing the notion, I think the core notion is a good one – adapting it. Sometimes I worry that these solutions have the opposite directionality. You know what we would call the Bangalore Tech bro solution, which is I’ve got a technology now I am looking for where will it fit, right? So I’ve got the UPI, I said, “Oh, that works.” Now where do I make it fit elsewhere? Efficiency was a big problem in the financial system. It is not my big problem in the health care system or it is a problem, but not tech-driven inefficiency, rather it is usage-driven inefficiency, which this can’t solve. In fact, it can exacerbate. So I think more comes from that reverse idea. I’m not saying that this therefore is necessarily bad. It just needs a deeper conversation to say, “What is the problem we are really solving?”

RB: [00:16:04] Actually, we are fully aligned on the principle. I think starting from the market failure that we are solving for rather than rationalising the innovation by articulating a market failure that it might solve, I think is very, very important. I think in privacy, information security is extremely critical. I think one of the principles that I’ve you know, that Tara Sharma once talked about with Aadhaar is what is the minimum information we need to collect about this person? Identifying this person is very different from how the common government collects data. That’s a lot. But I think respecting that in the design and ensuring that we design for it not just now, but on an ongoing basis, because it’s not a design issue, it’s a governance issue as well as you go along. I think that’s completely aligned as well, and third is it is not an ‘or’ you know, yesterday I was talking to somebody that sometimes we do something and say, I’ve done enough, now I don’t have to do other things. So to assume that I’ve built this digital infrastructure so I don’t have to actually sample water, I don’t have to invest in care is something that we have to guard against. It’s an ‘and’, and not an ‘or’ and looking at the other parts of the system I think is extremely critical.

NM: [00:17:05] If resources are limited may be an ‘or’ but against the technology and for the sewerage water collection.

RB: [00:17:13] I think that is in some sense the promise that we have to see happen for health. We think the open question is, is this helping us do more with less or is it taking away resources onto something else or other parts of it get starved, which is exactly what happened with TB and COVID, and I think I remember having a conversation with you then about diagonal systems, which is that the more verticalised systems they get, we actually have to make choices between TB versus COVID and not the other way around. But as this helps us do more with less, and one of the promises – and again I’m an optimist on this – are there ways in which this improves observability and that observability sparks innovation that we cannot imagine today where it might not have been designed to solve a problem, but the observability enables us to drive momentum around solving that problem more effectively. I think that’s what we have to see. But on the principles itself, I think we are aligned, and in my mind, that’s a paradigm I’m getting used to where we say, “Hey, here is a problem. I’m going to do a set of things to solve.” The problem here is an effective system. I’m laying bricks in the system. At some point, these bricks add up to find a way to create greater velocity to solve the larger problem.

NM: [00:18:24] In a way, what you said is right. If we imagine this system as you articulated it earlier, which is how can I meet the needs of the central unit and therefore this is the data collection platform for it, you know, I know surveillance, I know this, that and the other. But if we switch that around to say, “No, no, no, it’s not about that. How can I enable the fourth stage community health worker to become a more powerful provider of the latest and best protocolized service?” It’s not a conversation I’m hearing often. Often times what you get is “I’m collecting data.” For what? Monitoring, because the presumption is Semashko style. Everybody is a crook. Everybody is staying at home. I need to make sure that I monitor what they do instead of saying, “No, no, they’re trying to do the best they can. How can my tech – first and foremost, forget my needs – solve their needs? How can they get better?” And clearly the same tech may be important, maybe useful, but is a very different picture of what it tries to do.

RB: [00:19:23] I think that’s at the core of a lot of what we’re discussing. And we discussed it in the first part as well. As long as we see this as a command and control model, I think anything we create can serve that purpose and not this unlocking human spirit or human potential problem, which is what we have to solve for care-based work, which is what health care really is and not policing. I want to come to insurance and you touched upon insurance, but largely the question of financing. I think personally for me and I’ve had a chance to work on insurance as a problem, I’ve worked on financing as well, it is a vexing question to say “Who’s going to pay for all of this, and not just for today, because we’ve made some ambitious plans as a government to enable insurance for all.” – which is a good thing. But do we have the fiscal capacity to pay for it as our disease burden continues to grow? Part of the answer is what we spoke about in our first part to strengthen primary care, strengthen public health, and reduce the incidence. But you’ve spent a lot of time on the banking side and on the health care side. How do you see the financing question overall and what are some of the vexing realities or challenges that we have to overcome?

NM: [00:20:24] See, I feel in the insurance conversation, we have to go away from the words and go back to the first principle, the ₹2,000 we spoke about earlier, not the PMJAY ₹20, that is insurance, just not called insurance. It’s called free health care. So actually, we already have insurance, universal insurance. It’s called the free government paid health care system. Why are these conversations not coming together? Why is this being seen as the solution somehow that is not being seen as a solution when in fact, as somebody who works in the insurance finance literature actually think that’s a better idea because it brings financing and health care together rather than the financial instrument insurance which separates the two. In fact, this is a notion called managed care, which is now gradually becoming the dominant force, which was always our government health system, which is always our ESIS. So we are taking a good idea – instead of making that better because there are challenges with it – they’re not saying “No, no, we’ll go back to an older 1960s idea, reincarnate that and call that insurance.” We are going in a somewhat dangerous direction because if today look at one of the challenges in the US costs have gone out of control, where did it start? It started with John F. Kennedy’s assassination. He was very passionate about insurance for the poor and for the old. Lyndon Johnson took over as the US president, then won the election and he felt he needed to honour that implicit commitment John F. Kennedy had made, post his assassination. John F. Kennedy himself might not have pushed it that hard, but Lyndon Johnson – very quickly and there was a huge sympathy – he got Medicare, Medicaid passed into law in 1965. What was the error? The error was they adopted the Blue Cross Blue Shield fee for service, RSBY-PMJAY style design, which is unlinked to health outcomes. It is a pure financial product, and Mark Pauly wrote a nice paper in 1968 saying if you make a good free at the point of service, consumption is going to go through the roof. If you make a good that price is supply inelastic. No matter how low it is today, its supply is going to go through the roof. This is the US health system. This is what we have started, fortunately for us, it is underfunded, so we don’t really have the danger yet. So in some ways, my fear is we’ve taken a good idea, which is a government-run, integrated system, managed care provider, financier coming together. Somehow not try to fix that using purchaser provider split strategic purchasing. We’ve launched this new thing, which we started, you know, under the previous administration as RSBY, and we think that’s our answer. And as an outside observer, I’m puzzled by that. I don’t quite see it. I see that as the idea we got away from and the world is now recognising that it is not going to solve anything, it’s going to make things worse.

RB: [00:23:21] You know, it sort of sets the incentives in a way that does not align for everyone in this picture. You know, the doctor has a different incentive. The insurer has a different incentive and the patient has a completely different incentive.

NM: [00:23:32] In the 1980s, 1990s, somewhere near 80% of US market was indemnity style insurance. Today, 97% of the US market is managed care. This is illegal in India because we are in the 80s still where we are thinking insurance, healthcare is the same as fire, the same as property, the same as, you know, something else – that is healthcare. Not understanding that there is a very fluid conversation. Most health systems are now integrating.

RB: [00:24:07] But I also want to point out that managed health systems in India is given state capacity and ESIS as an example grows under spend, the money is there, but the transaction friction is at the hospital level.

NM: [00:24:18] So now that we get into a somewhat more sophisticated conversation, right, managed care now has evolved with the work of Eindhoven and others to manage competition. So Israel, for example, doesn’t have a single managed care provider. It has four. You choose which one. It puts enormous pressure, so there’s not one ESIS, four ESIS competing with each other. Now you say, “Whoa, if I don’t deliver the service, the guy can switch.” You can’t switch every day. There’s a restriction, but you can switch. In fact, many people don’t switch. The threat that I can switch is enough to build discipline. Within the government context, it was clear that managed care is a good idea, but you need a separate purchaser from a provider. Integrated provider, but the purchaser has to be separate because somebody has to hold the system accountable. Otherwise, benefits denial, which is the problem with this problem with state-run systems, becomes a huge issue. Insurance systems overprovide, managed care systems underprovide, ESIS and government health systems and managed care systems underprovide. It’s very easy for them to say you’re healthy. I don’t need any treatment. Then I die tomorrow of a heart attack.

NM: [00:25:31] Now, fortunately for us, the world outside has seen this a long time ago, but somehow we have gone back even further and picked up ideas that are even older and made them current ideas for us. Instead of saying, “No, no, actually the world did that”, it has already evolved to something else. In fact, Adam Wagstaff published a very nice paper to say this is an emerging Asian consensus. Even a hard-core left-wing country like Vietnam has switched. Turkey switched, Thailand switched. They all realised that if a government-run system integrated, and beautifully managed, has to work, you need to give accountability to it with trust. I give you money per capita for five years, but now I need to see results from it. I don’t know how many gloves you will use, how many syringes – your call – you want to build primary care, you want to build hospital-based primary care, you decide. But I will hold you accountable every so often with population-level outcomes. That’s the purchaser. This is government, that is government. Both are operating at an arm’s length basis.

RB: [00:26:35] For some of our listeners, we’ve made a jump to managed care and they might not fully understand or they might understand it differently as to what managed care is, so I want to – and even for my own understanding – articulate what managed care is, and I would like you to build on it. Today, the way it works is that there is a certain facility that is responsible for ensuring that good health care is provided. There’s another facility that sort of gains from paying for your health care, and then there is no one particularly taking responsibility for your good health in some sense. And these are all three different outcomes and not no one is particularly interested in integrating it together. A managed care model is where all these three in some sense converge, which is that there is an entity that says your good health is my commitment or my outcome that I’m accountable for. So I have the incentives financially to ensure that you’re in good health because we are not in good health, and if you incur diseases, I have to cover those costs. So I provide you the financial cover you need, but I gain by improving your health and reducing your health care burden. What we are discussing is if you can have 4 or 5 providers who can offer this, we enable the agency of the individual to then choose the right provider, but we also ensure that they get the care so that overall they get quality care and the fiscal discipline of providing health to people is actually managed. Would that be a fair description?

NM: [00:27:48] I think that would be a fair thing to kind of organise it a little bit. The first welfare theorem, which is when we spoke about market failure, one of its implications because you’re using the Adam Smith invisible hand, is that the price is a sufficient statistic. A C-section price is not a sufficient statistic. A price is not a sufficient statistic. A managed care price could be a sufficient statistic because now I’m saying to you, give me ₹10,000 a year. I will keep you. Well, I can take four people who made the claim and say, “Ah, I know what feeling well looks like, this guy did a better job.” I don’t need to get into – “Is the C-section the right thing for me to do?” Well, that’s your problem. I don’t understand it. I know what feeling well looks like, I can now let the market operate in a private system. In a government system – can’t have an external market. So now Mrs. Thatcher’s phrase – I build internal market. I build a purchaser and a provider. And I say, “Now I will let you work at an arm’s length basis because now it’s an internal market, but again, with sufficient statistics.” So that’s a nice transition from piece by piece. It’s like saying have a steering wheel price, I have a starter price, I have a carburettor price. Those are sufficient statistics. They are not. It’s the car that’s the sufficient statistic. Otherwise, you sold me pieces. I bought the best of everything. Now, I can’t make a car. It doesn’t fit my needs.

RB: [00:29:12] Even just to take the car analogy forward, it’s not just that I’m selling you a car, you might as well say that, “Hey, listen, all costs of running your car breakdowns, servicing is my problem.” So when you take accountability for the experience rather than just the utility that I provide.

NM: [00:29:26] And this is why sometimes efficiency is not the right conversation to have, because ultimately I’m interested in outcomes. It’s possible some quality may need inefficiency. You know, I may need to spend more time with you to get you to take the medicine, then I wouldn’t be efficient, but it may still be the best thing to do from an outcome perspective.

RB: [00:29:56] This sort of brings us to the conversation on strategic purchasing which the purchaser-provider. It’s not been, from a capacity state capacity point of view, a greater strength. I think there is much to be done. How do we make strategic purchasing more effective, especially in conversations?

NM: [00:30:12] I mean, where I would have hoped and if you go back to the report of 2011, the idea that we had put in there is that the national authority the state had, the authority is not running a separate scheme by itself. It is the purchaser. It’s not getting a small allocation of money from the government to run a private scheme. All government money is going to it, and it’s buying only from the public sector because this admission of failure that the public sector can’t do it, which is why now need to take some of that money and make it possible for people to go to the private sector too premature. I’ve never given the public sector the opportunity to perform. By building the accountability structures. So I would disagree that we don’t have experience. We have a lot of experience. We just are looking somewhere else. We think, “Oh my God, we need to create a purchase.” In fact, one of the ideas that I am pleased that this government has done relative to the design, is they’ve built trust, which has meant internal capacity. The good thing is, as it happens everywhere, some trusts have done well. Look at the Karnataka Trust. It’s done very well. It can teach the others. It can become the model to say, “This is how it’s done, guys.”

NM: [00:31:22] Now you convert that to say, “I built this capacity using a small amount of money.” But that was really only an experiment, a pilot. Now I’ve learnt I’ve done a 25 year pilot plus everything is many years now. I’m confident I know how to purchase care. I now switch this thing to say, “Forget the private sector for a minute, I’m going to give you all my state budget for health care. Now you buy from the public sector. I’m going to cut off all budgets.” – and Thailand did this. It needs political will. Doctors will go on strike. Nurses will go on strike. There’ll be all kinds of mess. But if you manage it well, people will see the opportunity. They’ll see the opportunity of more public funding going into the government system. Today you’re taking public funding and instead of building hospitals in Chhattisgarh, where there are huge gaps, you’re saying, “I’ll give it to insurance with the hope that some private provider will go to Bastar and build a hospital.” That’s not happening. Instead of that, tell the public sector I will give you all the money. In fact, more money, and I will trust you. The civil surgeon will no longer be an administrator. You know, just managing how many gloves and how many syringes he or she will be responsible for the district’s health outcomes.

NM: [00:32:31] If you read the report that Dr. Shreenath Reddy and his team – I was also one of them – we had the foresight to imagine that system at that time. We didn’t have the capacity necessarily, but with the SHA and the NHA, and I’m very excited about ABDM, therefore less about what it can do with the whole system, rather what can it do to the government system? I need to put gatekeeping in Kerala primary care. Too few people go to primary care. Too many go to the hospital. When I went to Thailand after they had done this exercise, I met the director of one of the districts where this was being done. I asked him, “So what is the single biggest change you saw post this implementation?” It was called the 30 baht scheme at that point. But essentially there are all these elements. The occupancy was 120% earlier – District hospital, the minute this scheme came, I realised I had the same amount of money. I cut occupancy 60% because a lot of patients that were there actually could have been dealt with very nicely in primary care. Why were they sitting here? Why were all these people in the queue? You go to any government hospital today, queues stretching for miles – you go to AIIMS, but you talk to 90% of the people they’re in the wrong place. He put a nurse, triage nurse, main gate 32nd question. Put in an auto, and sent back to primary care. Today, I can’t do that because I don’t have the primary care. Underutilised – there’s a very nice study by Sunil Nandraj and others. Utilisations as low as one patient a week, one patient a month beautifully equipped facility with nobody there when you can manage 200 patients a day and not just sitting in the clinic outreach. So think we have the capacity already. We just need to say let me think system-wide, let me not get caught up in “Ministry of Health will not listen, this will not listen.” And maybe that’s a central problem. At the state level, oftentimes the PMJAY and the Health department are both being run out of the health department. How can you make this work? How can you take, for example, a state like Goa and make it happen there for a while? They are ready, they are primed, they have the money, they have everything else. Take a state like Delhi, make it happen there. It works. Take it to Kerala. If it works, take it to Bihar.

RB: [00:34:30] One question which is on detail. I constantly think about what is the right unit for India’s size. There’s a lot to be done at the state and the state health association – and we have it, so it’s useful to leverage it. How do you look at the District as a unit?

NM: [00:34:42] So I would say let us not ossify the definition. Sikkim is 700,000 people. The whole of Bardhaman as the district think is, you know, 3 or 4 million. I think that’s the whole idea of high trust systems – I don’t legislate that, I build. In order for these things to work, you need a minimum population size of 10,000 plus. Given India may be a million plus, beyond that, there is no real benefit to size. You might say, “Okay, I don’t really care about district or state or anything else. How about a million people?” But then you might say, well, actually that’s also not a good metric because in Mizoram, in, you know, the million people is the whole state, but it’s so far apart. So there may be a geographic definition, may say no more than 100 square miles. So I would say let that be a dynamic endogenous definition rather than, say, district or village or block.

RB: [00:35:37] That’s an important thing because, again, we have this notion of top-down deciding many things and leaving it open, I think is standing to the principle. We talked about the public sector and strengthening it. What is our relationship with the private sector when it comes to health care? And I know this is different from primary, secondary and tertiary. In one of the conversations I had with a set of secretaries in the past, one of the points they made is that the state’s ability to regulate health care is very poor. You know, for them to even engage there is, of course, a line of distrust against markets in general and markets. And health care has a higher level of distrust than many other sectors. But at the same time, it’s seen as a necessary player in the ecosystem and not as somebody who is sort of here to make money. How do you see the private sector participation? What will be an effective model for private sector participation?

NM: [00:36:25] I worry about the statement that you are hearing that we can’t regulate because already we are getting into punitive, already we are getting into a thinking. When we say accountable, we mean I need a stick to hit somebody with. Instead of saying, how can I enable, how can I recognise, for example, that in primary care, the distributed set of actors that are out there, NGOs, pharmacies, they’re all private sector, they can do a much better job than I can do? How can I enable them to do that? US, for example, has a system called the Federally Qualified Health Centre lookalike. The Federally Qualified Health Centre is paid for by the government. The lookalike is simply branded by the government. It’s not paid for by the government. Why is that helpful? Because the signals to the people at large. This is the high-quality centre tomorrow. This is a vaccine campaign, a campaign. It goes also via these lookalikes. Why not do that right In a way you’re not regulating? You are participating, you are not financing. See, that’s the thing, right? The only two tools that a civil servant today things I have got is the dunda (a stick) or the money. But no, the state has many more tools that it can facilitate. It can guide, it can set rules. You know, it’s a bunch of rules, voluntary rules. That’s a great core design idea because it allows you to self-select good people who want to participate in your growth. Now, the reality is that and you see this in Sri Lanka quite a bit.

NM: [00:37:55] Sri Lanka also has very high out-of-pocket expenditures. But what they discovered is very interesting. And the first time I heard about it, it was really quite miraculous. They discovered that actually, it’s one of the few countries in which they have progressive out-of-pocket expenditure, not regressive out-of-pocket expenditure, which means it’s the rich that are spending out of pocket. The poor are not. But the rich in India is a very fluid definition. If you look at the global middle-class definition, my understanding is only 3% of Indians are above the $10 a day benchmark, which means 97% of the people are not necessarily poor, but they’re not even middle class. Are they rich? Not really. And then if you now say even within the 3%, it’s really the 0.1% that is so rich that it doesn’t even need any insurance. Most people, when hit with a, you know, pancreatic cancer bill of 35 lakh rupees will go under. Right. They will not have the money. The super-rich would not bother, but most of us would would say, oh, my God, you know, my life is over. So in some ways now you were to say, okay, I have to think about all systems, including the government system, for all people. That’s where the notions of managed competition now start to come in to say, first of all, the reality, unfortunately, of India is while the government is 10-15% of healthcare, the corporate sector is 4% of healthcare.

NM: [00:39:18] What is the objective reality of healthcare? Disaggregated individual providers dealing directly with individuals? Problem with that is, which is the perfectly competitive market, right? But that produces market failure because now all the machine asymmetry that exists that makes the first welfare theorem not work plays out. This would be a terrific market if it was toothpaste we are talking about, but it’s a very bad market design. The ideal market design is oligopsony is dealing with oligopolies, which is concentrated buyers dealing with providers, not individuals, because then both sides have symmetry of information that they can deal with. With the oligopsony being more powerful than the oligopolies that the buyer who represents the consumer is more powerful than the seller. That’s the game theoretic industrial organisation design that you’re aspiring for. Can we start to build that in the private sector? At the moment it’s illegal, right? If tomorrow a large health system said, “Where is the opportunity?” I teach a course on this at the Indian School of Business. What is going on today? If I look at any hospital balance sheet, occupancy rates are gyrating from the sun to the moon, 20% some days, 80% some days. ARPOB, which is the average revenue per occupied bed, is very low. You built a fancy gold-plated facility. Now you are doing, you know, routine fracture treatment in there. They could have been done in primary care, but you got to fill up the bed somehow because the occupancy is so low, and one of the complaints that hospitals have is why their valuation is so low.

NM: [00:40:48] You know, these new Start-Up Health tech companies are showing valuation signs that hospitals are not able to see when they’ve got all the stuff. The problem is what they don’t see and what the market sees is this is an unwieldy, uncertain structure. There are lovely – New England Journal of Medicine is a whole series on this – ‘Why did the US switch from indemnity to managed care?’ Not for any social reason. Hospitals realised that unless they enter primary care, secondary care financing for that care, they will not build a stable population that they serve, which means they will have empty hospitals, they will have sick people, but empty hospitals. With this structure, they now start to grow. You know, the unfortunate reality is if look at the sum total of revenues of all corporate sector hospitals, listed hospitals, it is smaller at $6-7 billion total than the top line of a mid-sized US-based hospital called Henry Ford, which serves a part of Detroit aggregate on one side, single system on the other. Why is that? Because all of this uncertainty doesn’t make for good growth, doesn’t make for good leverage, was a lender would not lend to these people would say, I don’t know, tomorrow your revenues could be so low you won’t be able to service my debt. But without debt, I cannot grow. If you enabled it, and then you said to people that don’t just buy insurance companies as one separate play, now you run a single play.

NM: [00:42:16] Somebody may say, “Oh, but they’ll create monopolies.” But remember our discussion about multiple monopolies? That builds managed competition. Regulation is not a good phrase because regulation implies perfect information. I can’t have I can’t say, Oh, C-sections are bad, so more than 15% will blacklist you. But I’m a referral hospital. Only emergency cases come to me. I will have 100% C-sections. That’s good. You want me to do a 100% C-sections. So you can’t regulate using standard crude tools. You have to use regulation without the laws, you are designing for good outcomes. You’re changing incentives because the incentives are wrong. No matter what you do, by law, they will find a way. For example, you saw I don’t know if you remember the old movies. Well, you’re not that old, but remember it. Juhu Beach has this gold smuggling point, risk-taking Amitabh Bachchan running around with, you know, lots of cash. What was that issue? The price signal was so strong between local and global markets that it made all that risk-taking worthwhile. People died, but people die all the time. But here they made a lot of money while dying. You have to bring the two together. Regulation can simply be an enabler as a patina on top so you know that there is a level playing field there, but it has to follow the incentives. I really think it can be done because corporates are hungry to grow. They don’t know how to grow.

RB: [00:43:48] Having a business model where your customer acquisition is assured is the best.

NM: [00:43:52] Now the incentives are aligned because you already collected the premium. Now you’ll only place a person in a fancy facility for a liver transplant, not for a fracture. And in fact, you will try very hard to make sure they don’t. If you’re an old person, they don’t have a fall. They will require a fracture treatment because you work with them to say, well, my primary care person will show up at your home to make sure you have enough handlebars and you’re doing enough tai chi so that your balance is much better because if you fall, it’s going to cost me.

RB: [00:44:23] I want to step back and ask you, you’ve spent a lot of time thinking about this. Are there other questions for you where you feel the answer is still unclear, where you feel there is a need for greater dialogue, conversation, research that will make us answer them with a greater sense?

NM: [00:44:39] No, I mean, clearly, to me, the primary care conversation needs a lot more thinking. I know the family is an important piece of the puzzle. I know culture is important. I know pharmacies are important. I know it’s community. Health workers are important. But how does it all come together? What if we are not able to build managed care and integrate it? What if the government doesn’t do purchaser-provider split and none of that happens and us hapless Indians are left with the current market structure? Is there an opportunity for a digital public good that what the Chinese call the internet hospitals? Is it possible to imagine a new market structure in which, like the classic prisoner’s dilemma, everybody is acting in their private interests, yet we’ve got a digital public good that is ensuring systemic welfare. I don’t know the answer to that. I can see the pieces of the puzzle. But how will this work? This issue that you just raised about will I create another ESIS through this structure? I don’t know. Benefits denial is a huge problem. I might say will create managed competition. But what if the customer doesn’t benefit from managed competition? It’s like, you know, remember once a conversation in which the regulator told a bank that you have too high non-maintenance or minimum balance penalties. The banker told the regulator that, “Well, sir, the customer can always switch.” The regulator got very upset and said, “Do you know how sticky the banking relationship is?” That’s why you got into it in the first place. Now you’re telling me they can switch, they will not switch and you’re exploiting that. Maybe the same thing will happen in managed care. I don’t know.

NM: [00:46:08] I think the NHA, SHA machinery is great. But is it enough? I’m thinking hospitals can turn into managed care organisations, but if you go and look at an actual hospital, the ingrained culture is so strong to fill up beds, to do procedures, to be busy, will they really have the managerial bandwidth to say, actually an empty hospital is best for me today? A provider has no interest in well or not. They don’t even bother with that question. Can they switch or do I need a new entity? There’s a lot of lovely activity happening on the health-tech side. Instead of saying, “Oh, let’s dismiss all that, it’s all dysfunctional.” There is natural momentum there. Can we benefit from that? Can we create what I would call a supra digital public good that takes all the so-called digital public goods and puts them to good use to say, okay, this is what you want to do. Please go ahead and do that and I’m going to change the environment. Now that ensures that while you are doing what you think is right for you, we are benefiting in the system. Think. Because I am certainly afraid that even the modern ideas that came to us from 70s and 80s may not fit well for us. Now we are going to the even older ideas. But what if we should actually be embracing ideas of the 2040s of the 2050s and not really going back to the ideas of the 1990s? I don’t know what those ideas are. I don’t know where the opportunity is. I sense it at the back of my mind that there is something there but can’t quite hold it.

RB: [00:47:40] I look at the history of what we have done right in India. Every time we’ve leapfrogged it seems like the right thing to do. We never wanted better landlines. We just went to mobile phones and we looked at data very differently in this country, and for all our connectivity challenges, the lowest price per data has actually changed things for us. We’ve done things other countries have not done, and that’s where the big leaps have come. And so you’re right in sort of saying, where are we looking for? Is it looking back at what other countries have done 20 years ago, or are we looking forward to a reality that is ahead, which makes it even more fuzzier for us to know that this works? There is no proof for us to build on. And the other thing that struck me is this is a larger macroeconomic question. I worry when oligopolies form from industry control in India, because a look at airlines, for example, look at other places. Our ability to sustain thriving businesses because of underlying functional foundational issues in our market have often been very difficult. You know, there’s always a failure where it becomes a duopoly and it’s either this or that and stickiness is actually very true. We design for stickiness in every business and healthcare especially so because the customer is so unaware. 

NM: [00:48:51] There are natural monopolies. You can’t have two airports. So if I’ve got a big hospital in a location, it immediately sends a signal to everybody else that, “Guys, this is my territory, don’t come here.”

RB: [00:49:00] And it is not as easy as setting up a mobile store, etcetera. So I think there are structural issues that sort of fall in the way of making this.

NM: [00:49:09] I mean, you see this in banking. Why is it that in India the difference between the market risk-free rate and the savings and current account rate as much as 400 basis points when the worldwide number is four basis points, why is the market forces not forcing a state Bank of India and HDFC Bank and ICICI Bank to raise these rates to market rates even though they are allowed by law? Why is it still allowing the regulator to say current account will have no interest? Why? In theory there is a market. In theory, there are players, but they have de facto cartelised, even though there have not been any meetings in the classic sense of the word that you can point to. But how come they all have the same rate? All the big banks have maintained exactly the same rates. It’s a real challenge.

RB: [00:49:52] For businesses like this, barriers to entry are going to be very, very high for any new player to come in and say, “Listen, I’m going to solve this problem very differently” which again interrupts the market.

NM: [00:50:01] But there, I don’t know. That’s what I’m saying. That’s where the fuzziness is. Is there an opportunity for a player to come in and say, “I’m going to look at this very differently and how can our digital public goods effort enable it?” How can we reduce the cost of entry so that now, you know, for example, for a while at least, the fintech world was going in this direction where banking, bank accounts became a service. Now you could do a lot more innovation without being a bank and therefore not threatening systemic stability because you’re not a bank. Do a lot of innovation and take a lot of risks. That’s a beautiful opportunity because now you’ve taken the systemic structure and reduced entry barriers. Now, is that possible to do? But we want to make sure that the right people are coming in because you reduce entry barriers, all kinds of charlatans can enter.

RB: [00:50:48] Then it goes back to the question of regulation and saying “Then we’ll put controls on who comes in.” And that often is ineffective for market behaviour, and to go back to our ABDM conversation are there ways to reduce lock-in? For example, if your patient data is not locked with a hospital but it’s made available for transfer, what does that mean? If the acquisition of patient health records, not just the patient becomes easier, what does it mean for us to actually accelerate innovation? I think the emotion that I’m leaving with this conversation is genuine, cautious excitement. I feel like, like you said, the pieces are falling into place. There are things that are happening like the NHA SHA conversation, the ABDM piece, and in a certain way, what it’s enabling is the digital public good conversation et cetera. How do we tie it all together?

NM: [00:51:29] No fundamental barriers seem to be emerging. You can’t say don’t have money. You can’t say don’t have doctors because actually don’t need the doctors. I have the tech. I don’t have surveillance, but I have satellites. So everywhere there is a possibility that’s really where then the pressure comes back on people like me to say, okay, man, if you said you are the integration, you are the research, you are the this, that and the other, what is your plan? Don’t tell me what the problem is. You tell me what the solution is.

RB: [00:52:06] It’s interesting how we end up with the first point we made in Part one of this episode, which is the intellectual infrastructure that we need to be multidisciplinary by design, learn from different systems, put together something that’s administratively feasible, politically viable, I think is truly what we need, and I hope in some sense this conversation sort of contributes to that.

NM: [00:52:27] I’m sure it will because I think there are many people, politicians, and bureaucrats who genuinely want this to happen. Academics, intellectuals that are interested in this think we have always had the sense that, “Oh my God, we have too many problems, let’s not even go there.” Right? You know, it’s just impossible. As I get deeper into it, I don’t see that problem. I don’t see that thing to say, Oh, my God, this just can’t be. This is the wall I’m hitting that I simply cannot climb. I’m not seeing those walls. But how does that come together to build a house? I don’t know that yet. Right. But clearly, the possibilities are there. So I like your phraseology of, you know, cautious optimism or cautious excitement. I would not even say cautious because that has kind of a negative ring to it. I would say, you know, be thoughtful about it. Be careful about which market failure you talk about what are you trying to experiment with? Don’t go in a direction that, for example, the stickiness issue stickiness is good because now you hold me accountable because I stayed with you and did want to stay with you. Maybe not. If you allowed me, my animal spirits would want a new doctor. Every day is a very nice paper by Dr. Kapoor that shows that for a single episode of tuberculosis, a patient consults six providers. Don’t enable that. Right? Make it sticky. Make most health systems will allow you one change a year. You can’t switch every day. Then the record transfer ability may actually make things worse, right? You might say, No, no, that’s not what I wanted.

RB: [00:53:57] I wanna really, really thank you. I think it’s been a fascinating conversation. I feel we’ve covered a lot of ground and I would love at some point for us to come back and say we made these prescient observations. How much of it did actually come through? Thank you so much for your time.

NM: [00:54:11] No, thank you for having me. Certainly, I am hopeful that we build a larger community of thinkers, ideators, you know, who are interested in the same questions and are, you know, reaching out to you to say “We’d like to do something with it.” and I think that would be a terrific outcome from this discussion. Thank you.

Rathish Balakrishnan: [01:09:53] Thank you for joining us here on Decoding Impact. We hope you enjoyed this episode and the conversation with our expert. To learn more about the Knowledge Institute and our evidence based insights, follow us on LinkedIn, Twitter and Instagram and explore our content on our website, all linked in the description.

Outro: Thank you for joining us here on Decoding Impact. We hope you enjoyed this episode and the conversation with our expert. To learn more about Sattva Knowledge Institute and our evidence-based insights, follow us on LinkedIn, Twitter and Instagram and explore our content on our website, all linked in the description.