India’s development experts discuss the country’s economic challenges, from underperforming public banks to insufficient human capital investment, exploring pathways to inclusive growth.
We need one last serious attempt at manufacturing
Dr. Arvind Subramanian
Senior Fellow, Peterson Institute for International Economics
One of the striking things we discovered while researching this book is the consistently poor performance of public sector banks after nationalization. We calculated the return on assets for these banks spanning from 1970 to the present, using Warren Buffett’s benchmark of 1.3% as our starting point. We made it even more conservative by using 1% as the threshold for acceptable performance.
What we found is sobering. Over this 55-year period, public sector banks earned an average return of just 0.3%—well below our conservative benchmark. In comparison, private sector banks managed 0.8%. This isn’t a minor difference. If public sector banks had performed at the level of private banks, the central government could have spent 25% more on infrastructure every single year.
People often say this comparison is unfair because public sector banks carry additional burdens. But actually, the onerous regulations—priority sector lending, statutory liquidity ratios—these applied equally to both public and private banks. What we’re seeing is a fundamental performance gap that was rarely scrutinized or acted upon by policymakers.
The consequences extended beyond just returns. These banks were forced into what we call stealth taxation and state capitalism. From 1970 to 2023, their primary activities became holding government securities and lending to public sector enterprises. As a result, credit available to the private sector—the blue line in our data—went from 8% to just 25% of GDP. Compare this to China or Korea, where credit to the private sector was a much larger fraction. This severely constrained India’s ability to finance manufacturing and industrialization.
On fiscal policy, we encountered another surprise. During India’s period of highest growth—1990 to today—our average fiscal deficit exceeded 10% of GDP when you combine central and state governments. This is almost the highest among all comparator countries. We call this the Kamadhenu fiscal state, because like the mythical cow, it suckles all manner of constituencies.
Here’s the puzzle: why didn’t India experience repeated crises like Pakistan, Turkey, Brazil, or Argentina, countries with much lower deficits? We did pay a tremendous silent cost—our interest payments are among the highest in the developing world. Today, interest payments exceed defense expenditure. That’s how much this vulnerability has cost us.
The interesting quirk is that the socialist state before 1990 actually saw lower deficits despite rising expenditures. It’s only when the economy opened up after 1990 that the neoliberal state became more redistributive. The composition shifted from public investment to subsidies and transfers.
On fiscal federalism, every federation transfers resources from richer to poorer regions—that’s part of the national compact. But what’s problematic in India is that these transfers keep rising over time. The biggest contributors aren’t actually the southern states, despite all the noise. Gujarat, Maharashtra, and Haryana contribute more. But the sense that underperforming states continue to receive increasing transfers without incentives to improve—that’s what creates the political discontent we’re seeing.
Let me address the questions about the Modi mystique and his timidity on reforms. On the mystique, we need to analyze this over time. The origin was less surprising—there was genuine frustration with corruption and economic mismanagement under the previous government. I like to say there’s a straight line from Pranab Mukherjee’s handling of the economy to Modi’s rise.
On sustaining the mystique, to be fair, there were positive attributes. I worked closely with him, and despite our disagreements, there was genuine commitment to getting things done. The concern for macroeconomic stability, fiscal consolidation, low inflation—these were real priorities I can personally vouch for.
But regarding timidity on reforms, there are several factors. First, he’s always been better at implementation than policy formulation. Second, the premium on loyalty over expertise has been rising, which means losing talent. Third—and this is perhaps most damaging—the belief that we’re already growing at 7-8% removed the urgency for reform. When I pushed for banking cleanup or other reforms in 2015-16, the response was always: why the urgency when we’re already growing so fast?
On where India’s growth will come from, I remain convinced we need one last serious attempt at manufacturing. The services-led model can only provide opportunities for 1-2% of the labor force. Even if successful, it cannot be inclusive by definition. And AI is actually a double whammy—it threatens to replace exactly those tasks where we currently have comparative advantage.
China’s share of low-skill global exports is 45-50%. India’s is just 3-4%. We’ll never reach 20%, but there’s no reason we can’t go from 3% to 10-12%. That would provide a tremendous boost to inclusive growth, even accounting for automation.
Why India has so severely underinvested in its people?
Dr. Devesh Kapur
Star Foundation Professor of South Asian Studies, Johns Hopkins University
Let me continue on federalism, because the usual debate focuses on centralization between Delhi and the states. But what we show in the book is there’s far greater centralization within states—between tier two and tier three of government.
Look at the distribution of government employees across China, India, and the United States. In China and the US, most employees are at the central and local levels. India is the opposite—most are at the state level. This is why our urban local bodies perform so poorly. They’re severely underresourced in both staff and finances.
This matters because the average subnational population size in federal systems worldwide is usually under 10 million. The exceptions are China and India. But China has far greater decentralization to its third tier. So not only do we have much larger subnational units, but these units are much more centralized than any comparator country. This is why urban areas, which should be engines of growth, have so severely underperformed.
When you look at revenues from land taxes, India’s have been flat and are a fraction of China’s. That bigger source of local revenues allows Chinese urban bodies to invest much more in infrastructure that helps cities thrive.
On human capital, we asked two questions: why were outcomes poor in the first four decades after independence, and when things improved, why only on some parameters? India’s ranking on the Human Development Index was 63rd in 1950. By 1990, it had deteriorated to 79th. After Sarva Shiksha Abhiyan and other schemes, it’s improved to 69th—better, but still quite low.
The improvements differ across aspects. On infant mortality, we’ve moved from 76th to 63rd—Bangladesh does better than us. But on child stunting, the improvement has been much more marginal. Stunting rates remain exceedingly high.
Similarly with education. The percentage of women completing six years of schooling has steadily increased since the mid-1970s. But learning outcomes haven’t improved—they’ve actually declined. More women are going to school and staying there, but what they’re learning has deteriorated.
We focus on two explanations: lack of resources and policy priorities. For the first time, we’ve computed the actual financial cost of state-owned enterprises from 1970 to 2020. At the central level, losses were 0.5-1% of GDP annually. At the state level, they were even larger. When you look at the opportunity cost, the losses on state-owned enterprises exceeded entire state health expenditures. Health spending could have been doubled without these losses.
But there’s also the priority issue. In the mid-1960s, it wasn’t the richest states that spent most on primary education—it was Tamil Nadu, Kerala, and Andhra Pradesh. They were in the middle in per capita income but had the highest expenditure on primary education. This was a reflection of policy priorities.
If you insist on throwing money down the drain on loss-making state-owned enterprises year after year, you won’t have resources for necessary public goods. Part of the explanation for poor human capital outcomes lies in priorities. If priorities are wrong, even adequate resources won’t deliver results.
Regarding the Modi mystique, sometimes you do well not because you’re great but because your opponents are terrible. I don’t have to be very good if my opponent is even worse. Part of the success is because opposition leaders haven’t inspired credibility.
There’s also an interesting dynamic about being single in politics. In the West, being single is a liability. In developing countries, it’s a plus because the assumption is you won’t make money for your family. To the extent that most political parties are no different from family-owned businesses, being single helps.
On the timidity regarding reforms, the spectacular failure of farm laws set back reform attempts. Many changes actually had reasonable basis, but agriculture is a state subject. When the center intrudes without consultation or building consensus, backlash shouldn’t surprise anyone. They withdrew and became much more timid going forward.
We should thank President Trump for what he’s done, because now people are realizing the world is getting tougher. The global environment is deteriorating, and that’s spurring some action.
On the demographic dividend question, one puzzle we explore is why India has so severely underinvested in its people—the one thing the country has in excess. Even now in higher education, India severely overinvests relative to quality. Our gross enrollment ratio relative to per capita income is a complete outlier. Most countries achieved these ratios at incomes three, five, ten times higher.
We’ve focused on quantity and forgotten quality. Most higher education doesn’t have signaling value. We’re doing students a severe disservice—giving them credentials that raise expectations but virtually no skills for labor markets. They’ll be unwilling to work blue-collar jobs because they have degrees, but they lack skills for white-collar work.
The World Bank clearly played a role in the push on primary education during the 1980s. India’s IDA share was declining, and the Bank said lending wouldn’t continue without agreeing to primary education funding. The world conference on education showed India had by far the largest number of illiterates. Unless we’re shamed globally, the bureaucracy and political leadership won’t act. That was the impetus.
What’s worrying is the coming delimitation after the census.
Dr. Duvvuri Subbarao
Chairperson, Madras Institute of Development Studies
When Arvind told me two years ago about writing this book, I wondered how they’d capture the complexity and diversity of India’s story over 75 years into a comprehensive narrative. They’ve done an amazing job. The Financial Times chose this as one of the best books of 2025, and rightly so.
Let me give you three reasons you must read this book. First, it explains not just how things happened but why they happened the way they did. There are hundreds of books on India from specialist viewpoints—history, economics, political science. Very few combine disciplines and explain the why, backed by data and analysis.
Take the Kamadhenu fiscal state. Democracies are supposed to be ruled by majorities but in practice they’re coalitions of vested interests. This coalition militates against common interests. The state accommodates all manner of constituencies—everyone has a right to become one. Indian democracy is an equal opportunity vested interest creator and sustainer, and the fiscal state responds accordingly.
Second reason: they explain developments that run counter to conventional theory. Interstate disparity, for example. In theory and practice, there should be no divergence across states in large countries. Interregional disparity should decline over time due to free movement of people, goods, services, finance, and ideas. That’s what happened in many large countries—the secret to American success is free labor mobility.
But that hasn’t happened in India. Disparities have actually increased. Punjab and Haryana—one state until the mid-1970s, same institutions, culture, traditions, natural resources. Today Haryana is the richest state and Punjab has regressed. Why? West Bengal, once the frontier state and capital of colonial India, is now a laggard. Kerala has no agriculture, no manufacturing, no services, but the best human development indicators—its economy is more linked to the Gulf than to India.
Third reason: they take bold, unconventional stands. Someone told me you should never make a statement a reasonable person cannot disagree with. This book passes that test eminently. On disparity across states, for instance—yes, cross-subsidization happens in all federations. Karnataka and Tamil Nadu get much less than a rupee for every rupee they contribute. UP and Bihar get much more. That’s understandable for national unity.
But what does it do? It creates moral hazard. Lower levels of government lose the urge to raise their own resources. Thirty years ago in Andhra Pradesh, when NTR called newly elected Panchayat presidents, we advised him to tell them to raise their own taxes—it improves accountability. You know what they said? “You think we got elected to raise money? No, we got elected to spend money. You give us the money.”
Most fiscal experts say taxation powers have been taken from lower governments. That’s not true. There are enough avenues. They don’t want to because they want the pleasure of spending without the pain of taxation.
Three issues merit discussion. First, fiscal consolidation and macroeconomic stability. Markets haven’t punished India like Pakistan or Argentina, but we were punished in 1991 and nearly in 2013. We didn’t commit as much excess. But we cannot declare victory. Central government debt-to-GDP is still 60%—the FRBM committee said it must be 40%. Interest payments are the largest and fastest-growing expenditure item.
People say our debt-to-GDP is low compared to Japan at 250% or the US above 100%. But we can get into a debt trap at much lower ratios. Combined center-state debt-to-GDP is close to 80%. The 60-40 picture matters: center collects 60% of taxes but spends 40%; states collect 40% but spend 60%. States’ collective fiscal stance is as important as the center’s.
Second, freebies. In a poor society where millions struggle daily, some safety nets are necessary. But spending on freebies is crossing limits, especially because this is borrowed money we’re passing to our children. A loan should repay itself—house loans, education loans do. But borrowing to give freebies is unsustainable soft bribery. Every leader offering freebies is saying, “I’m sorry I cannot give you the dignity of a decent life and regular livelihood. Please make do with this.”
Why do people vote for freebies? It’s very rational when you’ve lost confidence in politicians. If credibility is low, it’s rational to take what you get today rather than promises for tomorrow.
Third, the divide between rich and poor states. Thirty years ago, The Economist predicted an east-west divide—the east that flourished on public investment would decline with market opening, while entrepreneurial western and southern states would benefit. That didn’t happen. What we’re seeing is west and south versus the rest.
Last year, chief ministers in Andhra Pradesh and Tamil Nadu urged people to have more children. I never thought I’d see that day—fifty years ago, everything we did was family planning. The subtle message isn’t really about having more children. It’s about losing out—population growth rates have declined in these states while remaining high elsewhere.
What’s worrying is the coming delimitation after the census. Southern and western states will lose parliament seats disproportionately. So they’re being asked to continue cross-subsidizing poorer states while losing political clout. It’s a double whammy. How we manage this will be a major challenge.
Three questions: What explains the Modi mystique? What explains his timidity about economic reforms when he’s had more political capital than any recent prime minister? And where will India’s growth come from? People say we’re growing at 8.2%, but over the long term, we need 8% consistently for 22 years to be a developed country by 2047. We’ve clocked 8%+ growth just three times in 35 years and never maintained it for more than two consecutive years. Is that possible?
Institutional strengthening is extremely important
Dr. Soumya Swaminathan
Chairperson, MS Swaminathan Research Foundation
Thank you for including me in this fascinating discussion. I must congratulate you both on the extensive research—this book is a wealth of data that will take time to digest fully. Obviously my interest focuses on human capital, where I’ve been very concerned about the current state and our priorities.
I don’t have the answer to why we didn’t invest more in primary education or healthcare historically. Most countries in our region and Southeast Asia that have done economically well did so by investing in primary education. India had a commitment from 1946 to provide healthcare to all, with primary healthcare as a component. But we ended up spending more on tertiary healthcare and building institutions. Obviously those are needed for people who are already sick, but we missed the preventive focus.
I often give the example of Thailand. In the late 1990s, a political party convinced by public health people and doctors working in communities put a scheme in their election manifesto—30 baht a year for universal health coverage. They made difficult decisions: they stopped all investment in building bigger hospitals for the next ten years and put everything into strengthening primary healthcare.
Today Thailand has one of the most impressive primary healthcare systems. I’ve seen it—from village-level peritoneal dialysis at home allowing people to live quality lives, to different levels of care and how referrals work. They invested in institutions.
One lesson is how to prioritize based on evidence and science, informed by experts, but ultimately requiring political push. Sometimes decisions need to be shocking. I’m sure it shocked the Thai public that there would be no tertiary institution development for ten years. You need that courage, and you need to track and monitor trends.
India has been lauded for quality data. We have good systems and institutions. But using that data to learn and make course corrections—that’s where we fall short. We do national family health surveys every five years. There’s enough opportunity to adjust. The National Health Mission launched in 2005 with community health workers tried to strengthen primary healthcare delivery. But we’re still far away.
One issue is that health falls in the concurrent list, making it a state responsibility. If it’s state responsibility, we need much more decentralization to state level. But I totally agree—what’s holding us back is lack of fiscal devolution to the third tier along with accountability and responsibility for service delivery.
Most Panchayats except Kerala don’t play much role in the social sector. They’re not monitoring health or education delivery, not creating demand, not planning based on needs. We had recent experience working with Panchayats across one district developing a model development plan. It was uphill because they’ve never taken it seriously. They have no expectations that even if they develop good plans, they’ll get resources.
It’s chicken and egg, but we can learn from other states. Why repeat mistakes for years? Kerala shows how to devolve power to Panchayats. Tamil Nadu has built a very good public health system—not that we can’t do better, but other states could learn. The centralized medical supply division, central procurement, saves so much money. Pool procurement increasing volumes significantly brings down prices.
Tata Memorial Hospital’s cancer grid program showed that by pooling requirements from small cancer hospitals nationwide, they reduced cancer drug costs by 85%. We have models that work. Learning between states on education outcomes is crucial.
Institutional strengthening is extremely important. Not just scheme-oriented or program-oriented approaches where we scrap schemes and launch new ones. We need very strong institutions—regulatory systems need strengthening. India is pharmacy of the world. To keep that reputation, our Central Drug Controller must be considered among the best globally.
WHO has ranking systems for regulatory agencies—we’re not at top tier yet. For people to have faith in safety and quality of our drugs and vaccines, we need strong science-based regulatory systems. Similarly food safety regulators. Today unhealthy food is a big health risk factor. Unless there are strong regulators implementing front-of-pack labeling and taking action on advertising—anybody can advertise anything today—we’ll continue seeing increases in unhealthy food impacting health and obesity.
Preventive care is crucial. You showed stunting rates in India. While life expectancy has improved significantly, healthy life expectancy hasn’t. How many years do we live with chronic illness? A significant amount. Average life expectancy is around 70-72, but healthy life expectancy is only about 60. That’s a huge burden on individuals, families, and the health system.
We’re not doing enough to prevent disease burden. Prevention means providing clean air, water, sanitation, nutritious diet, good housing—the social determinants of health. Analysis shows investing in prevention is cost-effective.
Our health expenditure is well below the 2.5% of GDP committed in the National Health Policy 2017. Otherwise out-of-pocket expenditures remain very high. When you have out-of-pocket expenditure, you’re driving people just above the poverty line into poverty. Health expenditures are the biggest driver of sudden catastrophic expenditure resulting in huge financial losses for families. You cannot provide ongoing quality care this way.
Investing in institutions that develop policies is essential. Thailand has an institute for health policy and health promotion. Very cleverly, they use taxes from tobacco, alcohol, and sugary beverages—sin taxes—and put them into institutions working toward health promotion. The state then adopts policies developed by experts.
What can we do now to prevent our demographic dividend from becoming demographic disaster? If we don’t focus on early childhood nutrition and education, we won’t have people to make a developed country. Did the World Bank and IMF have roles in what you describe—decisions we made or didn’t make? Were we influenced by these institutions in areas where we didn’t do what we should have?
