Is India’s demographic dividend over?
Insights on what India must prioritise now to convert its demographic window into sustained and inclusive economic growth
In this episode, journalist and author Puja Mehra speaks with economist Arjun Jayadev, Director at the Centre for the Study of the Indian Economy (CSIE) at Azim Premji University and co-author of a major new study on how India’s demographic dividend has actually contributed to economic growth across states. Jayadev explains why India’s per-capita growth has diverged sharply across regions, how GDP per capita can be decomposed into productivity, employment, and demographic effects, and what this reveals about the three distinct phases of India’s post-liberalisation growth story. He outlines why the 2004–2017 period saw East Asian–level productivity surges in states such as Tamil Nadu, Gujarat and Karnataka, why employment ratios simultaneously collapsed, and how the years since 2017 have been defined by a worrying fall in labour productivity despite a rise in employment.
Jayadev also highlights the structural risks ahead: young states like Bihar and Uttar Pradesh that are unable to generate high-productivity jobs; ageing states like Kerala and Himachal Pradesh that are losing demographic tailwinds; and the national challenge of absorbing millions into productive work before the demographic dividend fully fades. Drawing on granular state-level data and labour force trends, he argues that India is at risk of “growing old before it grows rich,” and that the next decade must focus on expanding high-productivity employment, enabling migration, and designing state-specific strategies rather than uniform national prescriptions. Tune in for insights on what India must prioritise now to convert its demographic window into sustained and inclusive economic growth.
NOTE: This transcript is done by a machine. Human eyes have gone through the script but there might still be errors in some of the text, so please refer to the audio in case you need to clarify any part. If you want to get in touch regarding any feedback, you can drop us a message on [email protected].
—
TRANSCRIPT
Puja Mehra: Arjun, thank you for coming to the show.
Arjun Jayadev: My pleasure, Puja.
Puja Mehra: Arjun, we hear so much about India's demographic dividend and the non-trivial risk of it getting squandered. We also hear about jobless growth and that India needs a new development strategy. But as laypeople, we have a very vague understanding of all of these things.
You have studied this in great detail, including in how the demographic dividend has played out in the states. I'd like you to help us understand what is going on with India's demographic dividend.
Arjun Jayadev: So, you know, let's start with kind of understanding what the demographic dividend means. So, the demographic dividend refers to a period when a country has a rising share of working-age adults relative to children and the elderly. Why it's important and why it's important for the questions of economic growth is that the structure can sort of mechanically boost per capita income because there are more people who can potentially work and fewer are dependent.
Also, you know, it's a source of demand when you have a large young population. And in most places which are going through this transition, you have a period of between, say, 30 and 50 years in which you have this bulge of the population going through before they become older. So that becomes a really major growth opportunity for any country.
And India is right in the middle, maybe slightly towards the right now, of its demographic dividend.
Puja Mehra: Right, yes. So you have studied using real data how this demographic dividend has played out over three different phases. And you have divided the source of growth in GDP of states into different components, only part of which is the demographic dividend.
So if you could, you know, help us understand that.
Arjun Jayadev: No, this is actually the heart of our paper, Puja. So thanks for asking about it, right. So one thing that I'm a very big fan of in macroeconomics is actually making sure that whatever theories and whatever ideas and understanding we have of the economy, that it's sort of disciplined by accounting.
And one very good and very simple way of accounting for GDP per capita growth is to try to understand how much of it is coming from productivity per person who's working, how much of it is coming because employment is doing well, and how much of it is because more people are of working age. In fact, you know, I know this is going to be audio, but I'm going to try this anyway. If you can actually mechanically and fully mathematically divide up GDP per capita into three components, the first component is GDP per working adult.
You multiply that by working adult divided by, that's people who are employed, divided by the working age population. And then you multiply that by the working age population as a fraction of the total population. You have these three different components, right.
And once you see that, you can actually see that you can completely decompose the growth in GDP per capita into these three components. So GDP per capita may be growing because every worker who's employed is actually producing more. That's the labour productivity element.
There's actually an increase in the number of people who are employed of the working age population. That might have to do with, let's say, rising female labour force participation, changing norms, finding better ways to actually find matching between workers and jobs. So that could be another source of GDP per capita growth.
And then finally, there's the demographic dividend, how many people are of working age. So what I wanted to do was try to understand India's story, actually India's sub-national story, using these three components, if you will, of GDP per capita growth. And so, Avinash Tripathi, who you know, and I spent a little bit of time looking at state national accounts and looking at employment and working age populations.
The first is, of course, from our standard accounts. The other two are from PLFS, Periodic Labour Force Survey post-2017. And before that, we used the NSS-EUS, so the NSS Employment and Unemployment Survey, to get a sense of the working age population and the employed as a percentage of the working age population.
Now, using this data, we saw three natural breakpoints. So we looked at the data from 1994 to 2023, so roughly a 30-year period. The period, let's say, post-liberalisation, where we've seen this growth spurt in India.
And we found something actually quite startling. So we really found these three different periods, between 1994 and 2004, where we saw the first kind of growth, that, you know, the speed up in growth. So the component that came from rising labour productivity was high at that period.
What didn't improve, really, or was roughly flat, was the employment as a percentage of the working age population. What does that mean? That means that you weren't seeing people transfer from the working age population into employment at a greater rate.
Now, you must remember, India has roughly about, between 60% and 50% employment to working age population. Other countries, like the US, for example, will have about 80%. So there's a lot of scope for it to go.
And it's primarily women who are not working, who you could actually absorb in the labour market. That didn't change much. We did see, nevertheless, a rise in the demographic dividends.
So about half a percentage point growth per year was being added, simply because we were having a larger working share population. So that period was interesting between 1994 and 2004. But then the next two periods, between 2004 and 2017, and 2017 and 2023, were even more interesting.
Puja Mehra: Before we get to the next two periods, you know, should we first understand that the three components, what was driving them? And, you know, why did we see the productivity growth? And we saw the demographic, but we didn't see so much, the absorption, really, of working age population into the labour force.
Arjun Jayadev: Yes. So these are the questions that one really has to ask, right? I mean, what we've done is simply an accounting.
What we didn't go into, because it's really a complex question, depending on the period, the state, and so on, is which particular component was either contributing or not contributing. And, of course, you could imagine there's many reasons. I mean, one of the biggest questions in economics is why labour productivity rises in some periods and not.
But I think what is really interesting in this kind of decomposition was how differently the periods actually vary. So that's sort of a non-answer to your question. We don't quite know why these periods saw, say, rising labour productivity, falling labour productivity.
We only know that it did happen. That's the critical point. And even just knowing which periods saw these different components move gives you a sense of where the macroeconomy is now, what was the periods in which actually we saw higher labour productivity.
So that really is the purpose of this paper.
Puja Mehra: Right. Then what were the next two phases looking like?
Arjun Jayadev: Okay. So between 2004 and 2017, it's a very interesting period because, again, the demographic dividend keeps adding. In about half a percentage point a year, you're seeing people entering the labour force.
But we had this period that everyone knows now, jobless growth. Not only were there not jobs, you actually see a decrease in the employment to working age ratio between 2004 and 2017. And this is quite stark.
What that's saying is, whatever you're getting positively in terms of the working age population entering, you're actually squandering because you're not able to employ them. And in fact, you will see that the negative effect of the employment, shall we say, deceleration overwhelmed in that period, the demographic dividend that we were getting. A really tragic kind of situation.
Nevertheless, as you'll remember, between 2004 until 2010, say, and then after again, 2013 and so on, you actually saw fairly kind of robust growth. So what does that mean? That means that in many states, the labour productivity actually shot up.
And this is something that I found extremely interesting because if you look at states like Gujarat, Tamil Nadu, Karnataka, Andhra Pradesh, they're seeing roughly 9% growth per worker, per employed worker. Now, this is actually East Asian levels, close to Chinese levels. So you have states which are doing phenomenally well in this period.
And a couple of these states also saw absorbing of labour as well in this particular period, which is to say that the employment to working age ratio is actually stable or rising. So some states did extremely well in this period. And that, in some sense, is the golden age of labour productivity in this period.
Now, 2017 is the year in which the employment to working age population, you know, that's the middle part that I was telling you, the second component, bottoms out. And ever since 2017, I suppose the good news is that the employment to the working age population has been rising. That's out of everyone who's entering the working age, they're actually being employed.
Now, this obviously adds to GDP per capita in a mechanical sense. I should also be very clear that in some states, the demographic dividend is coming to an end. Kerala is an example, you know, where now the working age population is slowly declining as a fraction of overall population.
So there's a drag, if you will, a demographic drag in those places. But what does this also mean in an accounting sense? Between after 2017 up to 2023, as you know, we'd already seen a productivity slowdown between, say, 2017 and 2020.
Then you had COVID, then you have recovery a little bit after COVID. Of course, the data only ends in 23, we're slightly better off now. But what does that mean?
Because employment is rising, working age population is also rising, but GDP per capita is not rising fast. It must mean in a mechanical sense that there's been a collapse in labour productivity. What that means is that as in a mechanical sense, each worker is producing much less than they did in the period between 2004 and 2017.
And this is across almost all the states. There's only two or three states in which it's actually been a positive. So here's the kind of paradox.
Finally, when you're seeing a rise in employment, and I can actually say that that's a little bit misleading, but as you see a rise in employment, you're seeing no great effect on GDP per capita, which suggests that the employment that's being generated is low productivity employment. And therefore, you're seeing actually a reduction in the productivity per employed worker.
Puja Mehra: So essentially, you're saying that perhaps the new jobs that are getting created are of the informal nature and not the very, you know, the financial services and the absorption of labour is not taking place in those high productivity sectors.
Arjun Jayadev: Yes, I think that's a given. And I've been working on this with my colleague, Anit Basole, who's also been on this show. And we are actually trying to look at or take a deep dive into the periodic labour force statistics.
And it's very clear that the addition to employment, which is coming primarily from added women, which is in some sense a very good thing, is coming in agriculture. So you have for the first time in independent India, I think, a period in which the fraction of the labour force, which is in agriculture has increased between 2020 and say 2023. Now I don't know about 24 and 25, this data will come out.
So clearly employment is not going into the say high productivity manufacturing. It's not that that isn't there, but it's a very small fraction of the kind of jobs that are being created. So you're absolutely right, Puja, that the employment that's being created is not what we'll call high productivity employment.
Puja Mehra: So this obviously gives us a lot of insights into what the policymakers role is going to be now. And for them to take note of these findings, because in some sense, what you're saying is that if we do not begin to do things which ensure absorption of labour into more high productive jobs, we might even begin to come to the phase where all of our demographic dividend comes to an end without us taking advantage of it. Sufficiently.
Arjun Jayadev: Definitely. So this is the main concern, I think, of people, right? So as you know, India has got this extraordinary, even divergence in state per capita GDP.
I think Bihar is about a sixth or a fifth of Karnataka, say. And so you have some states which are young states, Bihar, Uttar Pradesh, relatively young states, where obviously people are now entering the labour force. Within those states, you're not seeing high productivity employment take off.
The southern states and some other states, Haryana and others, which are sort of older states, they may be slowly, not immediately, but certainly some of the states already, but they're slowly coming to the end within the state of having young workers come in. So there may be migrant workers who are coming in and taking some of the kind of slack, as it were. But where is the big kind of bang for the buck?
It would be in those states to see actually high productivity employment in those states in which there are young workers. And that's, of course, Bihar, Uttar Pradesh, or places like that. Failing that, you have a big problem.
You're either going to force people to migrate, and that's going to have its own dynamics and struggles, or you're going to see a kind of thing in which a lot of people are stuck in extremely low productivity employment, therefore very, very tepid growth rates over time. So the challenge at the national level is going to be obviously to see how labour can be deployed into places in which there's high productivity employment. But that first requires us to get high productivity employment at a much larger scale than we have right now.
Puja Mehra: So what you're also saying is that the way our economy was growing for growth itself, and especially per capita growth, more of it was coming from these states which were able to generate high productivity jobs or where labour productivity was high. But because of their age profiles now changing, we can't rely nationally on those states. Because the demographic dividend that can now come from the states, those states are not able to produce these kinds of jobs, and that is the policymaker's challenge, really.
Arjun Jayadev: Yes, certainly. So imagine that we're all different countries. That would be a problem because in that sense, you know, let's say Kerala or Karnataka or Tamil Nadu, if we only got workers from within the state, that supply would be drying up entirely by now, right?
Young workers.
Puja Mehra: Kerala is going to behave a bit like Japan.
Arjun Jayadev: Yes, and remind me to come back to the Japanese question in a second. But as you said, what we are lacking in that sense is a growth strategy for those states in which the young are. I think this is not a big secret, but if we are not able to get there in, let's say, five, ten years, we do have this very severe problem because even those states are going to start getting older.
And the big challenge, right, for India is to grow richer before it grows older. And in that sense, we're in a race. This last six years has not suggested that we're winning the race.
In fact, in some sense, we're running to stand still. There's more people who are working but at very low productivity jobs, which is why also you see things like relatively low real wage growth. So yes, that is the challenge.
Puja Mehra: Right. So should we now look at the different states, the groups of states, who's doing what, and you know, Kerala being like Japan, maybe?
Arjun Jayadev: You know, part of the reason for doing this paper was actually to think about what's the difference between really older populations like Japan and younger populations like other places. It turns out that, you know, we all know that Japan's GDP per capita has been relatively sluggish. But if you look at GDP per employed person, which is the labour product, they've actually been doing very well.
India is in some sense the anti-Japan, and that's actually the more worrisome thing, right, that we should be getting, because of our year-on-year demographic dividend, we should be getting at least half a percentage point more in GDP per capita growth, and we're not, because in fact, our productivity is falling. So we're working exactly the opposite. For them, their working age population is falling, but the labour productivity continues to hold them up with the exact opposite.
Now, what are states like, right? So if over this entire period, I would say, you know, you can say Tamil Nadu, Karnataka, Andhra Pradesh, Maharashtra, Gujarat, the same as you would imagine, they've combined in some sense, relatively good employment absorption. I can't say relatively good, because that period 2004 to 2017 was a disastrous period across the country, right, in terms of the employment absorption.
And then you've got what you might call lagging states in the sense that you have young populations, but they are not being actually absorbed. So Uttar Pradesh, Bihar, Madhya Pradesh, Rajasthan, they have young populations, weak job creation, and therefore low productivity, and also collapse in productivity growth in the last eight years, for the reasons I've mentioned. And the kind of states which are entering ageing, places like Kerala, Himachal, Tamil Nadu, they no longer have demographic tailwinds.
And so they must rely solely on productivity improvements or importing labour from, you know, these other states. So that's the kind of division I would put. And other states are more in one bucket than another bucket.
Puja Mehra: And just to understand, again, you know, if you want to say it with numbers, where you said some of the states were actually behaving like the miracle years of the East Asian economies, when I was reading your paper, the numbers really caught my attention, really helped me understand what was going on.
Arjun Jayadev: I mean, this is a not well remarked upon thing. And actually, I was very surprised to see this. So between 2004 and 2017, if you look at some of the states and things, Gujarat, Andhra Pradesh, Karnataka, Tamil Nadu, Tamil Nadu and Gujarat most, output per employed worker in that period, if you take, you know, the growth in that, that was about nine, nine and a half percent.
That's extraordinary and not something that you actually imagine, right? That Indian states per employed worker were growing at nine, nine and a half percent. And that's for not as insubstantial period, that was for a 13 year period that you're actually seeing this kind of growth.
And obviously that has meant huge divergence between the states, right? I mean, that has been something that we've been noting. I mean, anyone who knows the Indian economy says, it's actually not convergence, maybe it's district level convergence, but state-wise there's divergence.
And it's driven by this extraordinary period between 2004 and 2017. We've known about the so-called dream run between 2004 and 2008, but the fact that these states continued all the way till 2017 tells you that, you know, at the level of states, you've had some extraordinary growth stories, you know, and maybe next few years, I would like to see economists spend a little bit of time, you know, we don't need to necessarily only talk about the East Asian miracle. There's the Tamil Nadu miracle, there is the Karnataka miracle, there is the Gujarat miracle.
All of these places have been doing extraordinarily well in terms of output per worker, per employed worker.
Puja Mehra: In fact, I want to recommend here, I just finished reading, you know, Dr. Divesh Kapoor and Arvind Subramanian have a new book out called The Sixth of Humanity, and they have a whole section on states and they go into details of how different states have been doing very different things and a lot of surprises coming out there. And maybe tying in with, you know, what you're saying, maybe some of the intuitive explanations are over there. So, my last question would be that, I know this is not a part of your paper, but, you know, what would be the kind of things you would like to see the policymaker do now?
Not only for the Latin states, but also, you know, for the Karnatakas and the Tamil Nadus?
Arjun Jayadev: It's a good question, because, you know, there's no kind of ready-made answer to this. I think what it suggests is how different the policies have to be for each state. We sometimes think about a national level policy, but really we are just many, many different countries in terms of our demographic structure, in terms of the productivity per worker.
So, what are the kind of things that you'd have to see? I mean, it seems the most obvious thing is trying to find and actually produce work that is remunerative and high productivity. Now, obviously, this is like asking for, you know, the golden egg, right?
Nobody knows how to do that, but it has to be concentrated. I think it has to be concentrated and a lot of attempt to try to see, actually, to bring jobs to the people where they are. One of the things that we had spoken about before this was, you know, given the kind of size of the country and the kind of differences in per capita income, in any other country where there was, say, let's say one language or, you know, something that could help, people would be migrating a lot more.
China, that's exactly what happened. You had huge migration back and forth for decades from low productivity regions to high productivity regions and so on. And while that's happening to some extent in India, it's nowhere near the level that you have in other countries.
Migration is much lower in India than you might otherwise expect. That means that there has to be in-situ kind of jobs created in those areas in which the demographics are, you know, would permit young people to come in. So you have to much more focused kind of incentives for productive work to go to these younger states.
Also to try to make sure there's actually a matching for workers that actually can find jobs. Everyone is trying this, but I think the urgency has become extremely clear. But if we don't do it over the next 10 years, we've lost out on the demographic dividend.
Just to put a number on this, in East Asia, they did a really good job of employing and bringing people into the workforce. For them, it was one and a half percentage points over 30 years, which was the demographic dividend every year. That's extraordinary.
We have squandered it because of our employment policies, right? We've had half a percentage point. If we had employed everyone, half a percentage point or 6.7 over 30 years compounds to an enormous amount.
Puja Mehra: You're talking about growth GDP per capita.
Arjun Jayadev: No, sorry, let me be a little clearer. The demographic dividend, if you will, the working age population as a percentage of the total population grows about half a percentage a year, half a percentage point a year. And that, if you're able to employ them, mechanically should add half a percentage point to GDP per capita.
In East Asia, it added over 30 years, about one and a half percentage points to GDP per capita. In India, it should have added half a percentage point, but because we weren't able to employ that, it added actually more or less zero in many of these places, right? So that has been our great squandered dividend over the last 10 years, 15 years.
Some of the states, the Southern states, are not going to get that back anymore. They have to become more open to migration from other states and so on, right? And obviously that will have its own kind of dynamics.
So thinking very carefully about migration policies is going to be something that is inevitable going forward. And I don't know how it's going to play out with centre state relations and so on, but if you're thinking nationally, that's the only kind of thing that can actually help give the demographics to the Southern states. So that's the two different elements I would think of.
But again, this is at a very kind of, if I could wave my wand and if this could happen.
Puja Mehra: And I suppose you're also saying that the lagging states cannot anymore postpone manufacturing-led employment. They have to focus on manufacturing.
Arjun Jayadev: I don't even think it's a question of postponing. The problem is that manufacturing is one way to absorb. I mean, it's not even manufacturing.
It's not even services. There's just not high productivity employment available. It doesn't have to be manufacturing.
It could also be services, but you don't see that at all. That's the biggest challenge. Now, who knows?
Eventually you could actually see people moving in there. There may be policies which will allow for more, shall we say, investment to happen with greater certainty. There's many kinds of things.
People have talked about this for decades, but if it doesn't happen in the next five, 10 years at massive scale, we're looking at a really dire situation for the country.
Puja Mehra: Yeah, I read about these things a lot, but your paper has sort of raised red flags that it hits you in a way that, you know, I wasn't seeing it before this.
Arjun Jayadev: Yeah, it hit me too. So, I mean, that's why I wrote it.
Puja Mehra: Thanks, Arjun.
Arjun Jayadev: Thanks very much, Puja.
