Rathin Roy on How India Slices the Fiscal Pie

SHRUTI RAJAGOPALAN: Welcome to Ideas of India, a podcast where we examine academic ideas that can propel India forward. My name is Shruti Rajagopalan. Today my guest is Rathin Roy, a Distinguished Fellow, Kautilya School of Public Policy, and Visiting Senior Fellow at ODI Global. His main research interests are political economy, public finance, and development economics and he also served on the Thirteenth Finance Commission.

Hi, Rathin. Welcome to the show. It’s such a pleasure to have you here.

RATHIN ROY: Hi, Shruti. Thank you. It’s a real pleasure to be on this show. I’ve heard a couple of podcasts, and I’m looking forward to this.

RAJAGOPALAN: You are one of my go-to public finance people in India, and I would really love to geek out on all fiscal matters with you. Before we go down that path, just for the nonpublic finance geeks, can you lay out the broad contours of how India’s, in particular, fiscal federalism system works, and how do we as a union raise, distribute, and spend the revenues that we collect?

India’s Fiscal Federalism System

ROY: Thank you. I will. “That’s a very good place to start,” to quote The Sound of Music. The first thing, Shruti, I remember when I was a chief economist at the 13th Finance Commission, I used to anchor writing the chairman’s speeches to the chief minister of the states we visited. I wrote my second draft speech, I think it was to Goa or Kerala. In that I said that the Indian federation will not countenance cross-subsidies by any of its components to any of its other components. These speeches would go to the law ministry to be vetted.

Bang, it came back saying India is not a federation. Big shock. When you use the word union—this was 15 years ago in my defense. I was in my early 40s. I never ever called India a federation after that. India is a union, and it’s a union of two things, of the states and of the central government. This is something that I think your audience will be interested in because it’s political science. Hey, you’re at Mason, so you don’t care. You do both. The states can be created and uncreated and divided and subdivided and merged by the central government at will. The moment they exist, they become constitutional entities that are separate from the center.

RAJAGOPALAN: Yes.

ROY: India is a union of states with a center that has the power to create states or alter them, when in situ, the states have powers quite similar to the states in a federation, to provincial units, the federation. Our public finance structure mirrors this quite admirably. In our public finance structure, we have a pool of taxes that is collected by the Union of India. Then we have taxes that are not collected by the Union of India, that’s collected by the states and by lower units. The taxes that are collected by the Union of India, because India is a union of states, are collected by the central government in that the machinery of the central government is used to collect these taxes.

When these taxes are collected, they are distributed to the states. These are the major taxes of India, especially now that we have a GST or a value-added tax. These include all forms of income tax: income tax, cooperation tax, taxes of profits, surcharges, wealth tax, et cetera. These include all taxes on imports and exports, taxes on international trade. Mr. Trump’s lovely tariffs would be part of this divisible pool were we to have a Mr. Trump in India. It includes all indirect taxes: taxes on domestic goods and services in the form of GST. 

Then you have other taxes like motor vehicle tax or property tax, which the center does not collect; the state collects. Now, this has been an evolution. I’m just giving you the current, the extant situation. Much like the fairy godmother in Cinderella, these taxes, when collected and disbursed, someone has to decide what proportion stays with the center in New Delhi, the capital, and what proportion goes to the states, and then what proportion each state gets. That is done by a Cinderella institution called the Finance Commission. I was just saying I was the chief economist to the 13th Finance Commission.

We now await the report of the 16th Finance Commission with some trepidation, which we will go into later. The Finance Commission comes into being as a full-fledged constitutional body and writes a report. That report then argues more or less successfully what the proportion going to the center of all taxes should be and what the proportion going to the states of all taxes should be. It recommends that currently that number is 42% for the center and 58% for the states. 

It also recommends then the formula of devolution by which each state share will be known. That is called the horizontal devolution. The formula typically includes at least population area and some measure of income distance,  depending on people’s vagaries, some ecological measures, et cetera, but these are the three main ones. 

On top of that, it is enjoined to take account of a state’s capacity to discharge its basic fiscal obligations on the account that you should not be discriminated against in terms of basics depending on which part of the country you live, rich or poor. The metric for that is to look at whether when a state runs a deficit, which is its taxes cannot pay for its expenditures, is that deficit something that the state is running because it can’t help doing so to provide the services, the minimum service it needs to provide, or is it doing it because it is fiscally flagrant. 

The Finance Commission takes a call on that and provides what is called a revenue deficit grant to the states which it feels deserve it to close their deficits. They then provide other grants as well. Then when you look at all the taxes collected, the Finance Commission’s recommendations about the grants take those away. What remains? The center share, which is 42%, and the state share, which is 58%. 

Then this report is submitted to the president of India and tabled on Parliament, and the day it is tabled to Parliament, like Cinderella’s coach, the Finance Commission disappears. It cannot be questioned, it cannot be commented upon, it vanishes. Its members, its secretariat is dissolved. The whole function of the implementation of the report shifts to the Ministry of Finance, government of India. Then five years later and every five years, the process begins again.

That, in a nutshell, is how federal fiscal arrangements are devised. Now, there is one final thing I need to add, which is in addition to this, of late, the Finance Commission has also stipulated terms and conditions that limit the ability of states to borrow, that place a ceiling on it. It can recommend ceilings of the ability of the central government to borrow. But those obviously—you’re familiar with this—who monitors the monitor? There is no enforcing or binding constraint on the central government other than self-imposed constraint. However, in the states, there is a formula, and the central government monitors that and allows them to borrow or not per a constitutional article. That is also recommended by the Finance Commission. 

Then finally, since the center does not have the administrative machinery to devolve money to things that involve grassroots spending, there are schemes that the center designs and then co-shares with the states or provides the state’s grants to implement those schemes. Most important of those being the National Rural Employment Guarantee Scheme, various large health and education schemes, and there are several others. That, in a nutshell, is the panoply of fiscal federalism in India.

ROY: Or rather, correctly, intergovernmental fiscal arrangements in India. I must remember what the law ministry said to me 

RAJAGOPALAN: Before I went back to graduate school in economics, I got my degree in law. That also means that you have to read the constitution. I have always said Union of India. This may be because the hangover of all the case law is so and so against Union of India. For me, even today, it’s harder to say central government than it is to say union government. I have to sometimes correct myself because people sometimes wonder what I’m talking about, especially in the United States. I don’t know if it’s a feature or a bug, but it is an interesting characteristic of India that most people don’t quite imbibe what it means to be a union of states.

ROY: What you’re saying is so interesting, because if you look a bit at the history, maybe to help, especially your overseas audience to understand, India came into power at the stroke of the hour of 1947 as a completely bizarre child, both partition and united at the same time. Because before that, you had British India, and you had a bunch of people who wore turbans and jewels called Maharajas and— 

RAJAGOPALAN: 562 of them joined the union of India.

ROY: 562 of them. Dozens of gun salutes and some of them like the one in Hyderabad, this character could actually issue his own coin, and he had his own emphasis, they had their own arm forces. They had to be integrated into the union. At the same time, India was partitioned into Pakistan, Bangladesh, and India. That partition had to be dealt with, which meant that essentially the first few years after India became independent, in fact, it became a republic, we had to integrate these princely states, and we had to make sense of British India in terms of provinces.

There was no question of countries federating to join the union. The union was created as a result of both partition and merger of political authority, and then that union had to create subnational units, which would express the diversity of India. For better or worse, and that is also changing now, we largely chose language as the basis to express this diversity as a common ground, not history and not necessarily even administrative convenience or culture but language. Therefore we have had to adopt this very clever, I think, Cinderella practice of defining the relationship between the central governments and the state governments.

RAJAGOPALAN: That’s very, very helpful. This history is important. There’s a reason I also don’t like the term central government because I think it’s too much of a colonial hangover. We forget that this is no longer a government that is reporting to a small group of people in Whitehall somewhere in London. There are different power centers. To me, that’s an important distinction between federalism and just decentralization, is that under federalism, once you create a state, however flimsily it might be done—sometimes through voice vote, as we saw in the case of Telangana—however quickly and flimsily it is done, once you create a state, there is more than one power center, which is a really crucial thing. 

Now, when it comes to finances, we don’t quite behave that way. When it comes to, say, expenditures, we recognize there’s more than one power center. When it comes to elections, we recognize there’s more than one power center. When it comes to devolving from state to local government, we recognize there’s more than one power center. When it just came to standard garden-variety revenue arrangements, from 1950 to, say, 1990, that 40-year period, it seemed like there was no real recognition that there is more than one power center.

Is this because there was a planning commission and a five-year plan and there were plan expenditures, and the planning commission was so much more important than the finance commission, which is trying to deal with multiple centers of power, or is it just that was the beginning of the republic, and we didn’t have much in terms of revenues or coffers in the treasury, and it had to be a hugely centralized unit to make sure you preserve the Union of India before you can move on to thinking about multiple nodes of power? Can you give us some thoughts on that?

ROY: That’s a very interesting question. I think what you’re saying is correct. When the republic was founded in 1950, principally because the major source of revenue in the British period, which actually accrued in large measure to the British provinces and also obviously directed to the princely states, was land revenue. Because there was no other game in town. There’s a bit of industry in Bombay and a bit of industry in Calcutta. Some people paid a bit of income tax, but these were pittances compared to the extractive raison ‘detre for why the British were there in the first place.

Other than trade and imperial exchange, which was land revenue, closely followed by import tariffs. Now with the abolition of land revenue, rightly so obviously because it was an egregious tax on a country that was largely agricultural and where poverty was largely centered in rural areas. It was unfair and incorrect to tax agriculture to provide services to everybody else.

You were left then with import duties, which by definition have to be centrally collected. Then you were left with the task of building up a system of domestic indirect taxation and domestic direct taxation which would go co-terminus with the task of the industrialization of India and the task of increasing growth and prosperity. Note I did not say reducing poverty. That’s a development aim. I prefer the word prosperity very much.

RAJAGOPALAN: Same.

ROY: Especially today. Because the task was to increase prosperity and then to bring into the public resources domain, a portion of that prosperity to invest in India. The task also was for the government to be able to access what it hoped would be a greater and greater proportion of savings as this prosperity increased to use that through borrowing to invest in India.

Taxes to provide better services in India and borrowing to invest in India, but that process as it evolved took a lot of time. I would say that until we nationalized the banks in 1969, the state’s ability to actually tap into decentralized savings to be able to borrow, of course, prudently was limited. But only the center could actually create the machinery to be able to borrow.

Income tax and corporation tax as late as the 1980s were less than a third of total taxation; most taxes continued to be indirect taxation. That in fact was devolved other than what was called the central excise, which was imposed on a few commodities, not a few, actually a range of commodities that rich people, the middle class, very small middle class, the top 10 odd percent of the population consumed. That was shared with the states, and the income tax the center collected.

These tax pools, it was very important in the interests of equity because they were so small to make sure that wherever you lived in the country, you paid the same tax on these things, and you paid the same tax therefore on your income, wherever you lived, and you paid the same tax on what you consumed in these so-called luxury things on which central excises existed wherever you lived.

That principle was, in fact, cemented as the Indian economy became more and more prosperous in the ’90s and 2000s—we’ll come into this later—and cemented because quite worryingly they began to develop a huge asymmetry in the trajectories of prosperity in India. Therefore the tax base, while it broadened for both corporate tax and income tax, it broadened in terms of the total number of people paying it and the amount of tax collected, but not in terms of where it was collected from geographically within the union.

Excise was quite a mess. The dog’s breakfast in the 2000s, for example, all services, the indirect taxes accrued to the central government, and all goods, indirect taxes accrued to the states, except some goods, which everybody taxed like booze and cigarettes, and aviation fuel. All this was merged in what was called the goods and services tax data, which we can go into later. That in a sense is a bit more devolved. The history of devolution to summarize in the 1950s and ’60s was we are faced largely with a small tax net. Therefore it is important to tax all elites across the country uniformly because nonelites were not paying much tax.

We also wanted everyone who saved, their savings to be the disposal of government and the private sector to borrow. Therefore you had uniform rates of taxation across the country. A Finance Commission had to decide how those taxes would be distributed from the center to the states. Post 1990, while prosperity increased, and tax collection became more, while the amount of tax collection increased on both direct and indirect taxes, the geographical notes where taxes were collected became very polarized.

Therefore it became very difficult to argue for multiple taxation. We saw a similar problem happening with the GST. All these reasons, we had maintained uniform tax rates across the country decided by New Delhi, but distribution of those taxes is decided by constitutional body.

RAJAGOPALAN: Distributions, one, it’s supposed to be decided in this constitutional way, but it’s also decided using politics and other mechanisms. We have excellent work by public finance folks like Stuti Khemani or Arulampalam who show that union government is more likely to favor, in fiscal transfers, state governments which belong to the same party or the same coalition as the union government, and so on.

One is the formula of distribution. Within that formula of distribution, it’s useful to remember that there’s a fair bit of discretion. It’s not just a counting of sheep exercise. Once you bring in that fair bit of discretion, there are other things and political motivations which also come into how you actually distribute the pie.

ROY: I would disagree. I think it’s a classic example of atheoretical spurious correlation— by way of the research you are talking about. Because the entire Finance Commission proceeds are actually formula-based. When we had a planning commission, there was something called the Gadgil formula, which was used for grants. I would say about 70% to 80% of the grants we give are nondiscretionary, and to the extent that you see any favoritism, it’s not coming through political party alignment.

Any case, that argument would be quasi-spurious all the way up to 1994, because the majority of states were with the same political party. Today, also the majority of states are with the same political party, but there is definitely a correlation when you look at the third leg. On taxes, it’s impossible, there’s a laid-out formula. On borrowing, it’s impossible, there’s a laid-out formula. On centrally sponsored schemes, let’s take the National Employment Guarantee program, they say 75% of it is paid by the center say and 25% by the states, I can’t remember the exact proportion now. It’s either 50-50 or 75-25.

Now what happens is, even if it’s 100% from the center, you give the money to a state, how does it work? This is administrative. Let’s say you give the money to Bihar, and you give the money to Tamil Nadu. Now, to get your next tranche of money, you have to give something called the utilization certificate, which shows that you have spent the first tranche for the purposes intended. Then there’s auditing and all that. It’s just the first step. Now, what happens is that our administratively weaker states are less capable of pushing money out than our administratively stronger states.

Anything you do which is actually formally equal between Bihar and Tamil Nadu, Tamil Nadu will always be able to at least spend the money normally reasonably more effectively and do the paperwork because they’re good at doing it. They have 98% literacy, Bihar does not. That is a consequence of uneven development. I do not think that’s a consequence of political preference. That problem you do have. Then there is a counter to it, which we’ll come to later also, there is a counter that actually benefits states in these federal arrangements for money.

There are many other domains where you could say this. You could plausibly argue this is the case of, let’s say, the postage stamp policy on steel and coal in the ’50s and ’60s. Some people argue that there is favoritism shown to states in terms of getting the Olympic Games or the allocation of sports money between one state and the other. That is the allocation of specific funds for specific projects. The fact that Gujarat is the prime minister state and Gujarat is making a noise about hosting the Olympics, there you can talk about political preference. I’m not saying it is. I’m saying you can talk about it.

Not in the fiscal realm of taxes or borrowing, that’s also formula-based, or even centrally sponsored schemes. You have this defect, but that’s not politically related. Now, if you have a government that is only put to power in poor states and not in rich states, obviously, but even the BJP is in power in Maharashtra and Gujarat. Maharashtra and Gujarat benefit, but so do Tamil Nadu, so do Kerala, so does Karnataka, and Bihar and UP get screwed because they don’t have state capacity. That’s a state capacity problem.

India’s Welfare State

RAJAGOPALAN: Two things happened post-liberalization: One, as the economy grew, and we got out of this whole shortages/planning system where everything was controlled and rationed, we had genuine economic growth, which now the numbers are very clear for about 25 to 30 years continuously. 

The second part of that was that the government’s revenue also grew. India became truly a welfare state only after, say, the mid-’90s, because until then it didn’t have that much money to redistribute. It didn’t have 500 welfare schemes and programs. There was some welfare being done under plan expenditure and so on. Really the modern welfare state comes after 1995. Now within this modern welfare state, you have this bizarre interlocking going on because of India’s federal structure.

In India, you have things like health and education, which are technically a state subject where you will have a centrally sponsored scheme that is supposed to spend in a given state both with central and state revenues and finally be administered by the local gram panchayat, which causes this bizarre level of confusion on who is actually accountable, both in terms of personnel, in terms of policy formulation, and so on.

Of course, the second by-product of that is this phenomenal uniformity, which makes no sense because Bihar is different from Tamil Nadu. Also, there are parts of Tamil Nadu that are different from other parts of Tamil Nadu. Can you walk us through where is this craziness coming from? This is very much a post-1995 craziness. It wasn’t baked into India’s federal structure. I don’t know how to untangle this anymore.

ROY: It is possibly a job best left for an economic historian. I have been trying to do that because, let’s start at the bottom, firstly, something very interesting. When I was appointed to the Fiscal Responsibility and Budget Management Commission in 2017, with all sorts of worthies, much more worthy than me, I asked an initial question to my fellow members. I said, “I think we need to understand from the prime minister, from government in general, what is their desired size of the Indian state.” By which I mean, if you look at the total spending of the center and states, it’s approximately 25% of GDP.

Is this just right? Is this too small? Is this too big? They looked at me like I was mad and said, “What do you mean?” I said, “Is the size of the Indian state occurring by default? I asked this question to you very seriously, because the size of the Indian state”—at least that was before COVID, and after COVID is going back, it jumps up obviously due to COVID and is going right down—“has remained constant since 1991.”

Now, that’s actually a political statement. Despite India’s huge increase in growth, per capita incomes, everything—I’m talking about the union as a whole, the size of the Indian states at the center, plus states—measured as the public expenditure GDP ratio, has been constant since 1991. Which means we haven’t tried to shrink the state, and we haven’t tried to expand it. Now, fact two in this, in this traverse from 1991, the size of the central government has shrunk, and the size of the state governments has increased.

Fact number three, when the size of the state governments has increased, it has increased thanks to various fiscal incentives they were provided, in a far more fiscally prudent manner than the center. So you see, if you look at the difference between revenues and expenditures, which is borrowing. I get revenue tax and nontax, I borrow, and I spend, and that must always be equal.

Now, when I borrow over my revenues, I can use that borrowing either to finance consumption expenditure, which is wages and salaries, pensions, subsidies, centrally sponsored schemes, or I can finance investment. Now, paradoxically, when we look at the deficit, we have something called the revenue deficit, which I think is very useful, which what Gordon Brown called the Golden Rule, which is, “Thou shalt run government current expenditure on the basis of the taxes that thy collects.”

The central government did that very effectively in ’81. From ’81, 4 paise in every rupee that the government was borrowing was spent on consumption expenditure. By 1995, that had risen to 65 paise in the rupee. By 2005, it was 72 paise in the rupee, and it’s hovered between 70 and 75 paise, till I was in government till late 2018, 2019. The first thing was the central government, which was supposed to invest in India’s growth and development, was no longer in the investing business. The states, on the other hand, most of them, mixed picture, but most of them stopped running a revenue deficit in the 2000-2010 decade.

The heavy investors became the states, and the heavy spenders on consumption expenditure became the center. Because the center did not have the wherewithal to do investment, which was its original task—it used to invest in the public sector, it used to invest in defense, and it was fiscally constrained in terms of its ability to collect taxes effectively to such an extent, in a paper I did for the Pay Commission when I was on it, I found that the only government of India department that had grown in terms of manpower and size since 1985 was the Ministry of Home Affairs.

RAJAGOPALAN: Yes, Home Affairs, because we have security personnel just exploding.

ROY: Yes. We have them personally because we have internal, external security problems. You wouldn’t think that if we went to the airport and saw the people surrounding a minister. 

RAJAGOPALAN: No, but you do feel that when you go to the airport, because I’m old enough to remember when plane hijackings used to be like a normal Tuesday. We didn’t have proper airport security, and then it became fully centralized. It’s not all crazy.

ROY: It’s not all crazy, but optically, it looks horrible when you have 35 people surrounding some minister. I appreciate that.

RAJAGOPALAN: That’s terrible.

ROY: We have real security problems. Those security problems, I’m afraid we can’t wish away. We have them on our borders. We have them in states. We have had insurgent movements, and we have, as you say, airport security, physical installation security. That has expanded, but nothing else. Now, the center has a quandary. It has no money to invest in development. This is my theory. I’m still working on it. I’m too old to bother establishing it. I’m postulating it. I have data to show that what I’m saying is plausible, and that’s enough for me.

The young ones can run the econometrics if they wish. Just imagine, so you have liberalized, you have gotten out of the License Permit Raj. You are sitting there in 1995, Union coalition government, there is Mandal on your head, there’s security issues, planes are getting hijacked. Then Vajpayee comes to power, he goes in for privatization. What do you go to the electorate with as a promise? We didn’t have Hindutva in those days to help people vote for you, bigotry was not on the table yet.

You have to go essentially on a development premise. There was no money for the center, for the union government to say that I have improved your lives. You can hardly go to someone saying I have improved security, kept our borders secure and run the railways properly, which was a major central government spending. Let me just go back a second, if I may. I’d say very briefly that the developing state from 1950 to 1970 was building up the public sector, it was a national security initiative, it was also moving into manufacturing and creating the foundations for modern growth initiative.

From 1970 to 1990, we were trying to get rid of poverty, Garibi Hatao, we were trying to improve the quality of life and reduce vulnerability. We were remarkably successful in those things, poverty reduced et cetera. 1990 we liberalized, and then you have the central government which is now what do I do? I cannot be a developing state anymore because now the kind of investments have to make to be a developing state such as South Korea made, I simply do not have the money, and my per capita income has not grown that fast. I become what I have called in my other writings a compensatory state.

I am so sorry, I, Rathin Roy, and people like me, we have prospered, but you 85% at the bottom, your lives are not that great anymore. Here’s a health scheme, here’s an education scheme. Slowly the business of the central government moved into compensating people for the failures of development, climax being now under Mr. Modi where 800 million people in a country which has almost no extreme poverty, it has it in pockets, but aggregately, 800 million of our people, implausible that all of them are in extreme poverty, are reliant on subsidies for something as basic as food.

Our population cannot feed itself at current levels of prosperity. Then the state has to compensate them for that and give them food. The state has to compensate people for not being able to take care of old people. Here’s a derisory old age pension, but it’s better than nothing so take it. I think what the center went into then in that last phase was this business of compensation. Now, what are the states doing in this period? The states also realized that if the center is going to play this compensation game, then the political credit that they can get for many of the things they do can only come from competitive compensation.

Let me speak of the south where I live now and take Tamil Nadu or Telangana, it is quite noteworthy that all these states, what they did in terms of either creating manufacturing hubs or creating IT hubs or as in Telangana, in Hyderabad creating medical and knowledge hubs, was provide stuff like land and fixed assets, provide a conducive environment, and they got out of the way of the private sector. In terms of the investments they were making, they were making those on a commercial basis, infrastructure, et cetera. Their main investments were also going into compensatory expenditure.

Now it could be argued that some of the states, for example, Tamil Nadu and Kerala, and Telangana, their investments in education has done more than compensate. They’ve actually improved the quality of the workforce, the quality of life, et cetera. That’s an open debate we can have, but the point is that therefore everyone was going into the business of spending the public finances of current expenditure.

The center withdrew entirely from making investments, and the states decided that the best way they could encourage private sector growth development, manufacturing base, whatever, was to get out of the business of investing themselves or co-investing. There was some talk of PPPs for 10 years. It’s largely fizzled out and move into, again, the business of investing in human development and not a bad thing. Unfortunately therefore in the language of politics, the fisc is related now to the language of compensation.

Chief Minister Revanth Reddy in Telangana one day argues, “I have no money left,” and he’s frank about it. He says that, “If I am going to raise the levels of public provisioning I have, I have no money left.” This is a struggle even the richest states are facing, which is from providing merit goods, they have now moved to providing merit goods in the form of compensation. This is very difficult to reverse unless you have an ideology of overall prosperity, which we lack. You can’t escape political economy, don’t you agree, Shruti? 

Unequal Growth

RAJAGOPALAN: Absolutely. In everything that you said, there is one very important underlying structural problem which you didn’t unpack much, and that is even though we grew and quite fast relative to other countries after liberalization, one, that growth was not geographically equal; it was quite uneven. Second, that growth was not across all sectors in the economy. It was very much centered on services; the fastest-growing exports for a long time were services.

Now, there are two things that happen with this kind of growth. One, services growth is typically much more mobile than, say, manufacturing and agriculture. It’s much harder to extract. There is no fixed land and machinery that can be seized [Hindi language] the way they used to do in the 1970s. There is no farm where you can come and seize land or harvest the crop, the way they did with indigo farmers once upon a time. Services are just much harder to extract from, and they’re much more mobile, which meant that they voted with their feet and went to the southern states, which happened to have more development, better human capital, better basic services, and just overall better capacity.

Now, this means that our ability, our political economy, immediately shifts. You’ve explained one half of that political economy beautifully in all your columns, where you’ve talked about this compensatory politics. Now, in places like Bihar, where we can’t grant the same opportunities and which didn’t grow as fast as Tamil Nadu and Telangana, and Maharashtra, you’ve got to give these guys something. Now, we’re promising all kinds of freebies. There are agriculturally dependent states where the DISCOMs are going out of business because free electricity is announced, or highly subsidized electricity, and so on. That’s one part of it.

The second part is that states like Tamil Nadu and Telangana also recognize that they can’t exactly extract from Infosys the same way, once upon a time, Uttar Pradesh could extract from its industries in Agra, which means they actually have to provide something. What they’ve ended up providing is private enclaves. We will give you land for a technology park. We will build a metro for the urban people. We will do slightly better airports and things like that.

These are the things that the urban rich care about. These are the things that people in Bangalore care about because people are flying in from California, so the airport better be amazing, there better be a six-lane highway. That’s how these guys have managed it. You’re absolutely right. We have no theory of what the state is supposed to do. Second, we have no theory of prosperity, fundamentally, in a way that the state can actually do that, and then there will be prosperity in the economy. Does this line up with how you’ve thought about the problem?

ROY: Yes, absolutely. It does. Especially the last part about prosperity. There is another federal puzzle we can come back to, it loops back. The federal puzzle, let me unpack in two parts. I spoke to you about the horizontal devolution earlier. The taxes, remember, are collected by the center. Normally, what would happen if you had services, even services, anything, where you have the highly paid people, it’s not so much the income they earn that you want to tax, or their corporate profits. You tax them. As you’re saying, there’s a limit to that. 

You hope that when they spend money, that money will result in an increase in your direct and indirect tax revenue. In India, going back to the beginning, our direct and indirect tax revenues, the distribution of these are decided by the Finance Commission. 

You’ve taken away 42% and given it to the center 58%. What is the major component in that horizontal devolution which impacts the devolution? There are three major components: population, area, and something which every Finance Commission designs differently, which I designed in the old one. I still think my formula was the best. Anyway, I would, wouldn’t I? It is called the inverse of income distance. It varies, depending on the commission, between 45% and 55% of the total devolution.

It broadly means, simplest form, the poorer the state, the more they get. The most complicated form for somebody who understands that it is the inverse of your GDP rank, the distance of your GDP rank from the richest state. The richest state at this time being Goa, so how far are you from Goa? We finessed that. That’s technocratic bullshit. I took Goa and Haryana, averaged them, so Goa would get some money. Don’t worry about that.

Broadly speaking, it is that. The poorest state in India is Bihar, and the second poorest is UP. They get more money, far more money than the richer states, the middle-grade states. Now, the second most important is population, which is the most populous states? Bihar and UP. In the end, the growth boost that Telangana is getting or Tamil Nadu is getting or even Andhra is getting, not Kerala. Kerala has a different problem. 

The growth boost that these three states get and the expected revenue return they would get is just going away. Now, the states, I remember this, when I was in the 13th Finance Commission, we went to every state, and I was amazed at the level of solidarity that the richer states had with the poorer states, and they never said, “Why are these people getting this money?” They said, “Well, we have poor pockets, too. Please take that into consideration when you frame your formula. Let me try to help them out with some grants and things.” Nobody said what people have been saying the last three or four years because essentially the polity has become so divisive.

You start with Hindu-Muslim bigotry, and then you move into caste bigotry. One day you will come to north-south bigotry. It’s going to happen. PTR, the ex-finance minister of Tamil Nadu, currently the IT minister, he said, “We are not complaining. We are asking after 60 years of the inverse of per capita income, that is the richer you are, the more tax revenue you get. For 60 years after that being the dominant feature of how much each state gets, why are these poorer states still even poorer than us?”

That is one dimension of the problem. Let me as an aside tell you what Kerala’s problem is, which Amartya Sen never understood because he doesn’t do public finance. Kerala’s problem is it has great human development and great everything. What Keralites do with that is they vanish to more salubrious times to make their money like Dubai. Now, they come back and build pretty little pink houses in Kerala. That doesn’t help the government of Kerala. They don’t get their revenues. They don’t spend on petrol and excise.

RAJAGOPALAN: It’s not a dynamic economy in any way because all the dynamism is happening outside by Malayalees somewhere else in some other country.

ROY: It’s pretty dynamic in terms of the services they provide, the health, the education, their culture—they’ve become like Austria a little bit. Cultural dynamism and all that.

RAJAGOPALAN: I literally mean this is highest human capital. Why isn’t it the hotbed of startups in India or tourism? That kind of dynamism.

ROY: Because India is not exactly the hotbed of startups in the world. Keralites, being the smartest guys here, go elsewhere. They’re all in London now. They’re doing extremely well forming startups and things. That’s a very interesting paradox. They’re smart people. They look at India, and they say, “I’m highly educated. I can do better elsewhere. Why will I sit here? I’ll go to Dubai, or I’ll go to London.” Increasingly now London or the US till Trump, all bets off now. That’s the Kerala problem. Now, that is one dimension of the problem. The second dimension of the problem, which I’ve explained to some southern chief ministers, is their own fault.

Maybe it’s because they’re part of the union. The problem is that, I’m just talking about the southern prosperous states now, includes Maharashtra, my own state. I asked some chief ministers who should go unnamed one day. I said, “Tell me something. You have the per capita income of Indonesia. Why is it I can still buy tea at 10 rupees in an informal stall in Chennai or Hyderabad?” They said, “What do you mean?” I said, “Well, you’re so prosperous now that you should only be drinking tea in a formal restaurant, a modest one for 40 or 50 rupees like you do in Jakarta or in Bali.

“This 10-rupee tea reflects the fact that you are signaling that you want to persist with a low-productivity, low-informality economy, and you want to continue having three servants, nine drivers, nobody buys the dishwasher. You can do that partly because you can get cheap labor from the rest of India. You’re not making that transition to a high-productivity economy.” That is where the prosperity story in the south has gone very badly wrong. It is not just about the fiscal devolution. That’s one part of the bargain. There may I also point out what I have said in some work I’ve done, that is a very Faustian bargain that the Union of India has done since 1991.

South of India, the peninsula, where we have seacoast, beaches, nice place. If you come to India, come here. The rest of it is far too polluted. The south of India imports both capital and labor from the north and the east. Both, not just labor. In compensation for that, the Faustian bargain is that there is a redistribution through the public resource purse to the north and east of India. This is now straining at the seams.

Now, logically, as you well know, there are only three solutions to this, but that is maybe we can leave for later on. There are three long-term solutions. This is complicated by a political problem. The point is, therefore, that we are now caught in the fisc in a Faustian bargain. The summary of all that I’m saying is even the more prosperous states being part of this Faustian bargain and being also complicit in being unable to deliver a vision of prosperity that is fit for purpose for states at $3,500 per capita are now suffering stagnation in both growth and prosperity across the country, the net result of which I’ll tell you in a sentence, fiscally. What proportion of the Indian population is eligible to pay income tax? 7%. North or south. What does that mean? That is very troubling to me. When I started my life in 1989, I got a college lectureship briefly. I was paid something. I paid income tax. Today, somebody getting that same job does not pay income tax. An IAS officer joining the government does not pay income tax because their salaries are too low to pay income tax.

They start at 70,000 rupees a month. You start paying tax at 1 lakh, at Rs. 100,000 a month. It’s frightening. Lecturers in colleges, university assistant professors do not pay income tax. Doctors joining government service do not pay income tax. 1 lakh rupees a month is only what 7% of the population earns, and it’s not a lot of money. There’s a very good reason for that, because drive by the house today in India with 1 lakh rupees a month, 100,000 rupees a month, you can’t.

The kinds of things that you’d expect a prosperous person in the middle class to be able to afford with their own money are simply unaffordable if you tax them as well, which is why the proportion of people paying income tax is shrinking. It has shrunk this year from 30% of the population to the last tax reform to 7%. This is north and south. This is across the union. This is happening in every state. Basically, there’s a complete abrogation of prosperity that is happening as a consequence of all the things I’ve mentioned. This is commonly affecting the north and the south. I agree with you. I hope I’ve added some extra [crosstalk].

Consumption Taxes

RAJAGOPALAN: As a consequence of this, what India has ended up relying on in a really enormous and ludicrous way is consumption taxes, right? This is your GST. Now, there are many problems with consumption taxes, but the biggest among them for me is that they’re extremely regressive, because poor people just spend a larger proportion of their income on consumption relative to rich people.

Now, poor daily wage laborer who eats a packet of Parle-G biscuits at lunch is paying 18% tax or 12%, whatever is the current GST. I think it used to be 18% on Parle-G biscuits, right? What has happened to Parle-G biscuits as a consequence? They become thinner and smaller because it’s hard to increase the price from 5 to 6 rupees to 12 to 15 rupees because that’s where the squeeze is. Now, this is troubling in a few ways.

One, because consumption taxes are in general captive audience, very difficult to avoid. Once you integrate it with VAT and GST and India’s general tax surveillance state, it becomes very difficult to avoid. Pretty much everyone is on the hook for some kind of consumption tax. Second, it is also the least connected in terms of feedback to providing actual resources, and infrastructure, and health, and human development, and all of those things.

Walk us through what was the thought process on GST and now what is the consequence? You worked with Dr. Vijay Kelkar in the 13th Finance Commission. Dr. Kelkar is one of, not in the current form of GST, but he was one of the people who was the architect of VAT and a GST, a single unified tax, which has had enormous benefits. It’s actually gotten rid of all the check posts in India. It’s made India a single trading zone. There are enormous benefits, but there are these bizarre problems.

ROY: Yes, it’s a very painful history. It’s a history of, I think, how do I put it politely, because I find it particularly painful, personally? There was such promise there and we have now ended up with what I’ll call a ghastly and extractive tax. The idea of GST is very simple. I’d say at the outset, even recognizing that it’s regressive, the idea of GST is keep the tax low because it is regressive. Do not worry about equity, by tinkering with rates in GST, and then use the returns from GST to make things progressive.

It’s like the Kaldor-Hicks principle in economics. Do not worry about income inequality at the beginning. I don’t necessarily agree with that, but let people make money in the population of 100 and then take money from them and redistribute it to a lump sum tax. I think the same thing applies with GST. I had worked out that a tax between 9% and 11% could go even up to 13%, would because revenue loss to both the center and the states.

Over time, what we call the grey area dynamic effects, which we have not had, and the reason for that is not the GST per se, but how it has been implemented in the end, would make up for it and you’d have to basically manage for five or six years. Then the growth momentum of the Indian economy would bring you back to what one would call revenue neutrality.

The civil servants in the Bhavans didn’t like that idea. Their telescopic foresight is somewhat short. They convinced their political masters, who really don’t understand these things, I don’t expect them to, that what you needed was some absurd thing called the revenue-neutral rate. Then you had economists who didn’t understand these things, I won’t name them, saying that the revenue-neutral rate means including as many things in GST as possible. Now, the very point of GST becoming efficient becomes efficient because GST was replacing what is called a cascading tax system. 

RAJAGOPALAN: Which was so bad that GST is amazing. I want to caveat that, like things were really bad.

ROY: It’s not because we have replaced one sort of devil with another.

RAJAGOPALAN: No, that’s the consequence of how we’ve administered our GST. We can still fix that.

ROY: The point is that if you’re going to replace the cascading—no, we can’t anymore. It’s ossified, I think. We have to start again. So, the cascading tax meant obviously that your effective rate of tax on the final seller of a commodity would be much higher than the excise duty on the commodity because there’d be taxes coming into the cost of production anyway. Cement plus steel plus glass into a house.

The point is, therefore, if you’re going to change that into a GST, you’re going to lose tax, because tax burden will go down. You want a revenue-neutral rate, which is you want as much tax as before, in which case I calculated the tax rate would have to be as much as 25/24%. Then some great economists started saying, “No, we have to include real estate in this. We have to include petrol.”

I said that you lose even more tax, because those things bring in a lot of revenue because they cascade and their rates are very high. For example, if I didn’t bring in automobiles into GST, assuming I’d one GST, which I don’t, that’s my problem, the GST, then I would lose money on automobiles. Then you have to wait. Did you send enough automobiles to the GST to make up for it? It takes time. No, we’ll have a revenue-neutral rate.

Then I won’t name who certain people in government did some fudging and talked about gold, and came up with the rates we have, which were poor rates. A GST council was formed to look at these rates every three months or so. The result of which is such a dog’s breakfast now, such a mess, that at one point, we had separate GST rates for sanitary towels and sanitary napkins. You have separate GST rates for hotels, charging less than 4,000 rupees and more than 4,000 or 9,000.

What I had advised a hotel as a consultant, I said, “Just charge them for the room 8,500 and charge them for a towel 300, charge them for food 200, charge them for reception services and don’t pay—” People react, respond to things. People are not dead consumers, with the result of which we have now so many rates that actually I’m starting to wonder—

RAJAGOPALAN: We have seven nonzero rates, and we have over 50 cesses.

ROY: Yes, then you’re back to cascading, except you’re horizontally cascading, not vertically cascading.

RAJAGOPALAN: No, it’s worse than that. It’s worse than the old cascading, and I’ll tell you why.

ROY: I agree. I agree.

RAJAGOPALAN: I wrote about this, and this is a very George Mason perspective. When you say that you’re going to have a different rate for rooms below 8,000 rupees and above 8,000 rupees, or different rates for sanitary towels versus sanitary pads or whatever, now, everyone wants to be characterized then as the lower rate. There is massive amount of lobbying. I actually did the crazy thing. I tried to understand the dog’s breakfast, right? Which was a nightmare.

I went through every single GST council meeting and I created a table of which ones had a rate change. Okay. We’ve had more than 600 rate changes since 2017. These rate changes, now everyone wants to be a lower rate. Now obviously, that’s not possible because you want to be revenue neutral. Anyway, we’ve had COVID and all these other problems. What ends up happening and I haven’t run the data to prove this, but I’d love if some graduate student is listening to work with me on this, at first glance, or rather my hunch when I look at the table of rate changes is highly organized, oligarchic monopoly sectors end up getting the lower rate.

Your beedis and agarbattis and things that are made by a large number of people, relatively small sellers who can’t easily collectivize and organize their rates increase, because net, you can’t give a freebie to everyone. Now we have seven nonzero rates, we have 50-plus cesses, and we have a GST council meeting, which tells me that a barfi is different from a chocolate barfi. If they get put together in a Diwali packet of assorted sweets, then that’s a third rate.

ROY: Please don’t go on. It actually hurts me physically to hear all this. It’s a sign of a state that has lost its way. I tell them that. I say, “What are you going to Delhi for? You’ve just trivialized the taxation process.”

RAJAGOPALAN: It’s a sign of a weak state, but what are we doing as economists? We’ve gone bananas.

ROY: Economists can’t do anything about this now because we already abdicated our responsibility when we accepted. NIPFP, we got into deep trouble for it. I had to face a lot of pressure from the Ministry of Finance for saying that the revenue utility is as high as it is. The problem with economists is that once they become Darbaris, courtiers, if they start telling the king what they want to hear, which he should never do, because history will laugh at you later, then you will have to bear the consequences.

Now, it’s too late. Say, imagine if defense experts told India, when we went to war, that we had 700 aircraft, we only had 400, we’d been a mess. Just common sense. Now, there was another problem here.

RAJAGOPALAN: Now, before we get into the second problem, I have one question for you. We keep talking about revenue-neutral rate, right? Why is that the target? Why is the rate that we want to hit not the prosperity-maximizing rate? What is the rate that will spur the maximum economic growth in that region? Why is revenue-neutral such an obsession? It’s making me crazy.

ROY: Very simple. You’re not an administrator. 

 Your typical administrator: what are their objectives? If you sit in a tax department, what are you told? “Collect more taxes.” If you are in the Health Ministry, you are told, “Vaccinate more people.” Those are the messages people get. Now, in a sophisticated polity, with a slightly less weak state than we have, there can be a discussion on how these fiscal instruments should be used. Fiscal prudence to attain target.

India, in that way, might have moved into liberalization, and people talk a lot. Fundamentally, it’s a crude, I always call it feudal-socialist, both state and society, feudal-socialist. These guys are like, “We must have revenue-neutral rate, because that’s what we know.” What happens if you don’t have a revenue-neutral rate? Some hard work because you’re going to have less taxes than you had yesterday, which means you’re going to have to manage that.

Can you come up with a plan? If you ask the government of India today in 2024, about 2047, they will give you tomes, I’m sure. Everyone wants to talk about 2047. Ask them what India’s GDP growth will be in 2027 and what the structure of manufacturing, agriculture. They don’t know.

RAJAGOPALAN: We don’t know what our GDP growth was in 2019.

YV Reddy said in most other countries, the future is uncertain. In India, the past is also uncertain. 

ROY: That’s a very good point. 

It’s very simple. Basically, you said revenue-neutral because that’s what I do. Then you have access to the minister. You’re going to tell the minister, if you don’t have access to the minister, you’re going to tell the minister, “If you don’t have a revenue-neutral rate, how you’re going to lose the next election because you want to—”

RAJAGOPALAN: No, but hang on, Rathin, I want to interrupt you here. There’s a difference between revenue-neutrality as a goal and revenue-neutral rate as a target. For instance, if I charge 14% on an economy which is 100 rupees, I’m going to get 14 rupees. If I charge 7% on an economy which is 200 rupees, I will still get 14 rupees. Why is the goal not revenue neutrality in principle, as opposed to revenue-neutral rate as a number I need to hit as a percentage point?

ROY: Because they’re not willing to make that intertemporal tradeoff.

RAJAGOPALAN: In an economy that’s growing at 5, 6%, this is the country to do it in.

ROY: I took 7, but obviously, we’re not even capable of that. I took 7. 

RAJAGOPALAN: Once upon a time, we were growing faster.

ROY: I took 7%. At 7%, I calculated, not just me, other people had, I think Arbind Modi had, many others had, that by 2021, forget COVID, if everything had got business as usual, 7%, by 2020-2021, we would actually be exactly where you are. Obviously, what was happening was we were collecting 30 paisa in the rupee in 2014, that would go down to 20 paisa. When we implement GST. It’ll go up to 30 paisa, and then 40, and then 50, later on. They wanted their money now. They wanted their money in 2015, because I’m a secretary, I’ll retire in three years.

RAJAGOPALAN: That’s killing the goose that lays the golden egg.

ROY: That has always been India’s problem. I had a lot of problems with plank of planning we used to do once upon a time in this country. At least I had somebody who thought what would happen five to 10 years from now. There is no thinking on what is going to happen five years. Look at defense. We need 42 squadrons. We’ve been short of 10 squadrons for 15 years. No one’s thinking about where the next squadrons will come from, except when we have a war, everyone gets their nickels in a twist and runs around saying, “Sindoor, Sindoor,” I don’t know what they say. Three jets were shot down and we’d buy five. Indian policymaking is not professional in that sense.

When I say professional, I’m not blaming the politician at all. Both the courtier administrators and the courtier economists pandered to the politicians’ desire to behave like neoclassical economists and work exactly in one time frame, T0 to T1. Any attempt to go beyond that is resisted, including professionally. Even when I said that the revenue-neutral rate is best achieved, so you can have an alternative, I said, “Okay, keep the amount of things you put GST on limited.”

Then the revenue-neutral pain will be lower and then slowly keep including things till you come to 2021. No, if we exclude things and the revenue-neutral rate will go up. I said, “No, you’re not excluding, man. You are coming from a high-tax economy to a low-tax economy. The more things you exclude, the less tax you lose.” This is all professional stuff. Then, there was no argument for this.

NIPFP was the only organization saying this. In the end, we were completely ignored. Now, we have this dog’s breakfast and, good luck, Madam Finance Minister, with the mess you have created and the state governments who come and play trivial. I call it GST council meeting, not Trivial Pursuit. They have nothing more to do. They have no taxes to collect. They sit there and work on barfi and rasgulla and things like that. As usual, we have relapsed into trivia, even in the serious business of economic policy. Sorry, I think very strongly about this.

RAJAGOPALAN: No, no. So do I. I’m so depressed listening to you because you are exposing a problem. This is worse than what I had imagined. The way I had originally thought about the question was this was fundamentally a good idea, reasonably sound design, and completely botched up in its execution. I think what you’re telling me is that the execution is the design. Like, that integration is so missing in India that you’re never going to quite get the good design you were looking for, if we don’t have people who think about design and first principles and questions.

ROY: And prosperity.

RAJAGOPALAN: Prosperity.

ROY: If any government which said, “My goal is to get every bloody Indian not to compensate Indians—” I always say this. I’m not a communist. I’ve been saying this since my days at the UN that I think prosperity for me means what? I’m a regular guy. I go out there, I get a job, I get paid, right? Or I do a business, I get paid. There are five things I want to buy with that and say, just if I’m earning, forget poor, minimum-wage person or somebody earning three times the minimum wage, I want to buy food for my family. I want to buy clothes for myself. I want to build one house in my lifetime. I want to educate my family reasonably well, my kids, and I want some reasonable healthcare. How you finance it, public, private, forget. I want these five things and I want to spend my money to get them. 

None of these five things, as I’ve written about at this time are available. Agriculture is subsidized. Shirts for the poor, not my shirts, are made in Bangladesh and Vietnam. Housing, I’m reliant on charity called Awas Yojana, Indira Awas Yojana. Why? Why can’t they go to the market, get an EMI and build a house? That’s called prosperity and dignity. Healthcare, education, less said the better. Until we come up with somebody who says, “Hey, we’re going to be a prosperous country.” I believe that. Therefore, I’m going to have a tax system that is going to be something that is going to be a handmaiden to this path of prosperity.

Therefore, I’m going to have an expenditure system that is going to be a handmaiden to this path of prosperity, I focus everything I do on that prosperity agenda, then I would have a better tax system and a better fisc. The fisc is political, Shruti, as you know. The purposes to which it’s put depends on the purposes to which the state and society more generally desires the fisc to be used.

As long as the fisc is going to be not used for prosperity, you’re going to have criminalized GST, you’re going to have a compensatory state, and you’re going to have the paradox of prosperous South Indian states unable to harvest their prosperity and continue to sell tea at 10 rupees a glass when they should be like Indonesia, and at least selling tea and coffee in a formal establishment at 30 rupees.

RAJAGOPALAN: Yes. So, it’s a consequence of the fact that about 900 million of the 1.45 billion people in India are still living at wages which are very low and are living in a world where economic growth hasn’t picked up at all. In fact, productivity in agriculture and other areas is dropping. The only way to provide them some dignity so that you don’t have a revolution and people invading rich people’s homes and taking over factories the way you did in communist Russia is to have some of this compensatory welfare redistribution mechanism that we’ve come up with.

Public Finance in India

Now. Let’s take a step back for a minute, right? Now I’m going to go crazy George Mason public finance on you, right? Let’s start with, you know the Buchanan, Vincent Ostrom, Dick Wagner, that kind of view. You want government and the taxation of the government only to be at the level of the externality. This was something that was hammered into us. National defense, you obviously want the union government to do. At the state level, you may want broader consumption taxes or something like that, and at the local level is the garbage pickup. Now you call that local taxes, user fees, whatever your night watchman state, you choose your formula.

Within this, there is some room for intergovernmental transfers, but the lesser the better, because intergovernmental transfers fundamentally sever the linkage between voters and taxpayers. That’s the tightest link we have, given that the fisc is entirely political. If I take this mindset and go all the way with it, what it tells me is that most of the taxes should be collected locally because most of the things that people want are local schools, garbage to be picked up, public sanitation, some basic healthcare, which can very easily be provided locally. I’m not talking about the fancy AIIMS or something like that. I’m talking about the first interface of most people’s healthcare system.

The next stage you can talk about the state government, which does most of the expenditure, and then the union government should only stick to six or seven things. These are like defense and space, and maybe if it’s getting into education, it could be AIIMS and IITs or setting up DRDO or something like that, but really, that’s the realm it should operate in. This is exactly upside down of what we’ve done for 75 years. Question one is, is this a vision that we can have for India in the next 100 years? We couldn’t have it in the last 75 years. If we wish to have it for India in the next 100years, how do we get closer to that?

ROY: As a grad student, during my PhD, I did a critique of public choice. One problem I had with it was that, while I learned a lot from it, I must say, as I’ve always said to you, is that they looked at tax and expenditure, and borrowing separately. Buchanan was far too tax-focused. The very interesting,  Samrudha Surana, I think he’s in your school.

RAJAGOPALAN: Yes.

ROY: He and I have had a conversation about a very interesting exchange between Solow and Buchanan. I knew about them because I happened to be in Cambridge then. I saw it. It was actually partly about this, that Buchanan was entirely very focused on taxation and the consumer, not on expenditure. In India, what you have is the principal instrument of public finance is not taxation. It’s actually borrowing.

Therefore, automatically or the political economy is very simple, which by the way, public choice does acknowledge the problem, as of course you know, the problem with internal debt is not that it creates debts for the future or that neoclassical garbage. It’s fundamentally regressive. That’s why I don’t like too much of it, because you borrow from rich people and then, as I’ve said in something I wrote, you pay them for the privilege of developing their own country.

I’m a rich guy, I put money in the bank. You take my money, you build a road and I drive on that road. Nice, and I get 7% interest. Lovely, love it. Borrowing is highly income regressive. It’s as regressive as taxation. One must be careful with it. Given the fact that we borrow from the rich, actually, the expectations of the rich with respect to the public finances is less on the allocation function and much less distribution function, which is essentially a task of national security in three dimensions, I’ll explain that in a minute, and more about the stabilization function.

For the rich to earn returns in India, which are not high risk, given the undeveloped state of equities markets and the lack of investment opportunities now for a long time more generally, most banks invest in what? They invest in fixed deposits. Fixed deposits go to whom? The governments of India, the state and central governments. Some mistake people made when Rangarajan set this limit of 9% of GDP for the fiscal deficit or 7%, with financial savings at 9%, if the government of India is going to borrow 7%, there’s only 2% left for the private sector to invest in.

Indeed, that is what is happening. Almost two-thirds of our savings in the domestic system are going to the government. Borrowing is the prime mover of spending and that complicates what you’re saying in terms of hierarchies, because borrowing is entirely driven by the central government, sovereign borrowing, because the states are not sovereign. The ultimate liability for borrowing exists with the center. Therefore, the center has asymmetric power. Not the taxation so much.

Because the center has asymmetric powers into borrowing and then you add the political incentive, if the center is not able to manage defense, foreign exchange, and railways effectively, and has no big national scheme to run, then you have a problem. Now how do you get around that in developed countries, developed federations such as Germany? There you get around it in two ways. The center provides infrastructure, the center provides opportunities for export-led growth and the federal government, and the federal government most important of all three provides you overall social security and national health spine.

None of these our central government has the state capacity to do. What Mahatma Gandhi said about Western civilization¾My view on India’s export potential¾“it would be a very good idea.” I’ve been hearing that since I was five. I’m now 59, and our share of world exports is still zilch. Please export and then we’ll see. Not yet. Now, when it comes to a national health system or a national spine, the central government is simply not able to do that and therefore what it does is it again showboats.

I’ll open an AIIMS in your neighborhood, I’ll open a Y in your neighborhood, I’ll open a Z in your neighborhood. It does that because it’s political credit actually comes from providing welfare. There’s one public finance contract going on with the rich on the borrowing side and that is where money is spent on investment incentives et cetera et cetera, and the other public finance contract is going on with the poor.

Add to that the problem that our geographical development path has been so uneven and what I’m coming to the elephant in the room but the population growth has also been uneven that we are now coming up with a situation where the prosperous parts of India in the south and the west, particularly the south, face a hemorrhaging of political power if the delimitation were to be done as it ought to be on the basis of population, whereas they have the resources.

The center is having to navigate and only the center can navigate this. Now, I was saying this at the beginning. There are three forward-looking solutions to this, all of which are tectonic in terms of the political economy. The first is very simply that people from Bihar and UP migrate to the south. That’s voting with your feet, right? That’s the old thesis that should have happened, except that I think because the caste system that’s too long a conversation and many other things, when people from the north and the east migrate to the south and the west, they do not become of the south and the west.

They continue to hold both their vote and their political authority in the east and the north, and the south and the west don’t accept them as Malayalees or Tamilians. They’re still migrant labor. We use that phrase in the country. That’s not going to work. The second is that the south develops so fast compared to the north that it’s able to subsidize them and more and keep the union intact and not basically may have a representational voice, but does not have political voice because does not have the economic clout.

Some countries have survived well that way but no country in the world, in my reckoning, has seen a situation where the relatively less populous parts of the country are five times, not one time or two times, five times more prosperous than the relatively less prosperous parts of the country. But those prosperous parts of the country have three times the population of the economically more prosperous parts. High population-low prosperity, low population-high prosperity only happened 100 years ago in Yugoslavia. No country of federated units of diversity has survived this.

RAJAGOPALAN: At that time, Yugoslavia didn’t have one person one vote, which we do have which is important in this calculation.

ROY: Exactly. Very good point. If you look at the US, just for you US people, who cares what happens in South Dakota? They don’t count. The more populous parts of the country, well, let’s see what Trump does, are New York, California, Seattle, Illinois, et cetera. It’s okay for these people to subsidize the less prosperous. Same with Europe. Low populations, they belong the less prosperous places in Europe, Ireland, and Greece and whatnot.UK, France, Germany, these are the muscular parts. High prosperity, high income. 

In this condition, applying the public choice thesis of devolution is going to create, I think, the national security constraints to the reimagination of this, unless we are able to come up with a political settlement that promise is inclusive, prosperity and equalizing prosperity.

In that settlement, what you’re saying I think is both feasible and desirable, but until the settlement comes, we are not going to have the possibility of aligning our incentive structures in fiscal federalism with the very appropriately described incentives that Buchanan, Tullock, et cetera describe for governments to act, and that distribution between different units, which is also why very briefly in India the third tier is a complete fiscal chimera.

They’re not fiscal units. When people talk about loans to Bombay, I want to say Bombay is not a budgetary entity, Chennai is not a budgetary entity, and they will not be because they are sources of extraction of rents which are far too valuable to the state government to just give up as if they were communities.

RAJAGOPALAN: Okay. Lots of things to respond there. I’ll start first with the Buchanan critique and so on. I partially understand what you’re saying about the Buchanan-Solow debate, but I think there is enough work by, say, Buchanan and Dick Wagner and books like Democracy in Deficit, where they really do connect the tax and the revenue-raising question with the spending and deficit question and the debt question, right?

Because the issue with debt is not just that it’s regressive. It is that it’s intergenerational. The people who benefit today are not going to be the ones who pay things back if you have massive deficits, which you will then cover with debt or inflation or something else. This political business is both redistributive in the present, but it’s also redistributive intertemporally, which is exactly what democracy is able to do.

You can get votes today and have lots of spending, not raise taxes, not cut back on expenditure, on whether it’s welfare or whatever, wasteful expenditure, agricultural subsidies, fuel subsidies. We still do fertilizer subsidies. That kind of stuff. Now, the only way out of that conundrum, the way I think about it for a place like India is dramatic fiscal federalism. There really is no way out. When you talked about all the three problems that you mentioned, the only real way out is we reduce the reliance on intergovernmental transfers and we actually give the money back to the people.

Now, giving the money back to the people of Bihar can happen in two ways. Either we use the union government to take money from Tamil Nadu, and Kerala, and Telangana, and Maharashtra, and we hand it over to the state politicians of Bihar to then pass it further down. Another way of doing it is let the people of Bihar move to Tamil Nadu, which you told me is politically complicated, but it will be less politically complicated if they become people who contribute to the revenue of that state.

If their benefits and the revenue become portable, then suddenly, a Bihari is very valuable. I live in the Washington, DC area and Washington, DC and Virginia, and Maryland are fighting for my taxes all the time. They want to attract me both at the local level in terms of the county of Fairfax where I live, but also the state of Virginia where I pay sales tax and so on. Now, that doesn’t seem to happen because we’ve not made it easy to exit agriculture.

We’ve not made welfare benefits portable for the most part. I think now in food subsidies, it’s a little bit portable. A Bihari family can’t port their benefits or take whatever the government will give them and take it to Tamil Nadu and say, “Now, I’m going to work there. My benefits go with me and whatever I buy in terms of bidi at the tea stall and shirt and everything, on that I’ll pay state GST and all that will go to Tamil Nadu.” I think that is where the crunch is.

ROY: No, there are three problems with that argument. One, that the intergenerational argument doesn’t work with internal debt. As long as the rate of interest on debt is less than your nominal growth rate, you’re good. Our growth rates, fortunately, have been good enough nominally, 12-13% as opposed to a nominal interest rate of 6-6.5%.

I think what you’re saying about the migration is far more interesting. See, the problem there is that migrants from Bihar and UP to South India at this moment, this is something South India has to solve, are working in extremely low-end occupations. Therefore, it’s not a matter of the portability of their benefits. It’s a matter of their portability as citizens to a place.

Now, you living in Washington could live in Maryland and no one’s going to say, “You can’t live here. You can’t marry X, Y, Z from Maryland.” Whereas, take my example, my father’s family came from Bengal 150 years ago to Barak. Because we were Brahmins, there was portability because they understood Kashi, they understood all the places where we burn our dead, and we do our grandfather’s shraadh.

Now, a Telangana Yadav and a UP Yadav don’t intermarry. A Telangana Dalit and a UP Dalit—there are only two communities in India which are extremely prosperous, which intermarry and populate. If you look carefully, the IT hubs and all these modern global centers, those are Baniyas, who have a national community, and Brahmins. They’re only 8% of the population. These guys working in tea shops, restaurants, informal sector, they don’t bring the kind of prosperity to the state.

There is no prosperity argument in the south, to receive them and make them part in a declining fertility zone of their place. Instead of just saying learn Kannada, you have to incentivize people to learn Kannada and you can do that if they know that if I learn Kannada, I’m going to become an engineer. There are complex social reasons why that reception is not happening. If that reception happened, that problem is solved.

RAJAGOPALAN: The reception is happening at the business level. 

ROY: The macros don’t work.

RAJAGOPALAN: No, the macros don’t work for a different reason. Okay. So, there are two parts. One is the social reception and the economic reception. I think the economic reception doesn’t happen because the most local form of government, which is where all the congestion and the burden is, right? Someone will come and settle in tenements, or someone will come and increase the amount of garbage, or someone will come and increase the congestion on the streets. That person is not paying user fees or taxes or compensating that unit of government in any particular way. The economic benefit of having these people come remains at the business level. Construction companies love them because they get cheap labor. Saravana Bhavanloves them because they can work in the back.

ROY: Shruti, their consumption would multiply if they brought their families with them and they’re not bringing their families with them, a large reason why there’s no housing. It’s a very interesting experiment in Kerala. The Kerala government built migrant labor housing for families and then they came and learnt Malayalam. That is not happening at scale. That is not happening at scale across the south.

Let me put it this way, if the population of the north, if we had a long-term goal that should decline by 15% through migration and the population of the south should increase by 15% keeping the total population intact through migration, then what you’re saying is, again, then we need a long-term prosperity plan. You can do much more. You can set up a lot of industry in the north, et cetera. It’s not happening.

The very minimum, the short-term solution to what you’re talking about this divide, is to bring people here. I can give you a very interesting example of the socioeconomic constraints. These 200-rupee shirts that keep containing the import from Bangladesh and Vietnam, they’re made in Tirupur, Tamil Nadu and Vapi, South Gujarat. I’ve gone and met those people. I’ve said, “Hey, you’re making 3,000-rupee shirts in India. Why don’t you make these 200-, 300-rupee shirts? Go there.” There’s a polyester rayon problem. I said, “I’ll solve that. If I can highlight it, I can solve it.”

They said, “No, sir. Wage costs there.” I said, “Wage costs there are very low.” They said, “We don’t have skills.” I said, “All your shop floor workers are from Bihar and Jharkhand. Go there.” Then, basically, they said they were not comfortable operating in that environment. This is a classic public policy problem that states, and governments, and civil society, and society can solve. In India, what is happening is, I think, worse. Society is ossifying.

It is regressing into ghettos, which is making that outreach that would make—and I agree with you, that’s part of public choice, I agree with that, voting with your feet is the only long-term solution to the survival of this federation. Anything else is fraught with national security risks. There is no reception for this, either politically or within society. Neither is South Indian society ready to receive the Biharis, nor are Biharis ready to be received into South Indian society, unless they are from a caste level or income level which is so large that because prosperity in India is not inclusive, they become a minority, such as in Bangalore.

You will get some street fights, in Uber and all that. The majority of the Bangalore rich, Baniyas and Brahmins, don’t learn Kannada. There is no incentive for them to do so. The poor, paradoxically, do learn Kannada, because they have to be waiters, et cetera, but they are not contributors to the economy. We are caught in a problem which is much bigger than economics. Unless we have a national conversation on prosperity, we are not going to be able to solve it.

RAJAGOPALAN: That’s my last question for you. Does this require a national conversation or does this require the 250,000 local level governments that we have to slowly start becoming more capable, more powerful? There was a lot of devolution in Kerala from the state to the local government. We saw that happening in Tamil Nadu. Now, between after Kerala and Tamil Nadu, Karnataka started devolving more to the local level, right? Is this the only way out that, instead of having one big national conversation, we have a quarter of a million conversations happening parallelly, happening competitively. That’s the only stable equilibrium in the long run. When I say long run, I mean 50-year run.

ROY: 50 years, I don’t know, because in all these you are looking at, basically, there is one very simple Buchanan-compatible solution to all these, outcome of all this, which is the breakup of the Indian Union. If we don’t want that, we will want a national conversation. That national conversation will obviously, if it’s genuine, like any conversation—

RAJAGOPALAN: What breakup of the Indian Union, Rathin? People keep saying it. I don’t buy it. The people who we break away from, how will they defend themselves?

ROY: You don’t need to break up the Indian Union not to have internal secession and weakness. You don’t need to actually break away. You ask someone from Tamil Nadu this. We are going off topic here. They will say, “What do I care? What do I have to defend myself against? China? I’m not interested. Pakistan? What do I care about Kashmir? What do I defend myself? I don’t understand what you are saying.”

“I’m not worried about Chinese coming, rushing to the states. I can’t really deal with them. I’m perfect living in Tamil Nadu. We support it.” We don’t want these conversations to happen, because this union is a beautiful experiment, at least for me. I think it is very important for us to carry the union with us. There is no getting out of the national conversation precisely because of the complexity of this union, the languages we people speak. What you don’t want is local governments coming in and further ghettoization into the languages, regional identity.

It’s a large country. When you talk with the best in the world, the largest local government, look at what happened in Andhra and Telangana, you could call that decentralization, but it could also be viewed as something that is negative. To receive people who are not of you and like you, which is what Indian society is like and what the Indian Union is on an experiment about, you will need to have a national conversation.

That will not be a genuine conversation unless it is embedded in the north and in the south, in many local conversations. That part I agree. That neat distinction between national and local will not work here, because India was formed precisely because the centrifugal elements that we needed to instill required us to conceive—Nehru never called India a nation. Never. He always called India, in conclusion, a nation in the making. I think they’re still there. To make this nation, we need to always have a national conversation. That national conversation becomes meaningless unless you have conversations. Here, I agree with you. Conversations that are nontechnocratic and therefore local. 

Trouble is that because of Brahmanism and people like me, too many of us have simple technical fixes to these questions. Those technical fixes will not go and that political input, the social input that I’m requiring, which was there in Kerala, will not come unless you have conversations across India, in all the localities and regions of India. That has to be part of a broader national conversation and nationally imperative. That part I agree.

RAJAGOPALAN: Yes. One, I agree with you. One is we need to step away from technocratic solution. I think the second thing we need to step away from, and this is very uncomfortable for the technocratic classes, we need to step away from this obsession with uniformity, which is different from equality. We want everything to be uniform between the quarter of a million local governments in India, which is simply not possible. You can’t even have uniformity within Tamil Nadu, let alone between Tamil Nadu and Bihar.

I understand that there is going to be some ghettoized elements and some regressive elements in some local governments. There are going to be these Khap Panchayat-type problems, which exist even today with high level of centralized government. I think it also provides for a lot of experimentation. You’ll also have lots of things like Kerala, where you do invite outsiders and you do provide housing for them. My vote would be towards more experimentation and less obsession with uniformity. That experimentation will give us some rotten apples. I think I am personally comfortable living with them, but I know that’s an uncomfortable thing for most other people.

ROY: I’m all for experimentation. Again, we can’t do it technocratically. I need to see people in the local areas of Tamil Nadu, for example, saying that they want that experiment. I need to start the conversation between them and Biharis. They don’t care now.

RAJAGOPALAN: I’m saying let them at least start the conversation within their own state. Madras and Coimbatore and Trichy are not the same as other parts of the fishing coastal communities of Tamil Nadu, which are far poorer. We need that conversation first within Tamil Nadu.

ROY: Who’s going to have that conversation? You see that the incentive is not there, because metropolitan centers is where our extractive centers are. So, you could do this. You have to have an ideology of prosperity within which you can do this. This is something Charan Singh said a long time ago. He refused to call himself a Jat. Forget his successors. They have, as usual, defenestrate his legacy. He said, “I’m a farmer’s leader.” If we move to those categories, farmers, workers, fishermen, right? Consumers, women, these are categories that can speak across ethnolinguistic divides. The first start would be the ideology of doing that. Politics will not do it, because politics is circled around these divides.

Society will not do it because caste is circled around these divides. Unless we have a social movement at the very minimum that is able to do this and is able to come up with categories, where such conversations are possible, what you said is excellent, because the fisher person in Tamil Nadu, the fisher person in Maharashtra, and the fisher person in Odisha have a lot to talk about. So does a cremation ground worker. 

That is irrespective of how prosperous Tamil Nadu is, other than Bihar. What you don’t want is the cremation ground worker to be Bihari, and everybody else to be Tamilian from the Thevar caste and from Madras to be run by Tam-brahms. That is where we are now. That transition, which you are describing, I’m not against. I’m saying it has to be a political and social conversation. I don’t see the political leadership for such a conversation because our political rents are not divided that way, but one lives in hope.

RAJAGOPALAN: One lives in hope and the political rents are not divided that way. Going back to all the fiscal questions that you mentioned. The fisc is more deeply integrated into our life. I hope listeners at least get that out of this conversation, that it is not the separate, arithmetic counting the beans exercise. It is deeply entrenched in everything we do and how we speak to each other, both as citizens and governments. Thank you so much for doing this, Rathin. I love speaking with you. I always learn so much from your writings. This was such a pleasure.

ROY: I really enjoyed this conversation because it’s nice that people take these things seriously and think about them. The only message I want to send through your podcast for Indians listening is, for God’s sake, think about this and not how much tax you pay and how your local rubbish is collected. There’s a political economy behind it. If you don’t engage and participate in that, nothing is going to change. Then if nothing is going to change, you might as well go to Dubai. Good luck and goodbye. Cheers, Shruti. Thanks.