Guided story
India's R&D Puzzle: Tiny Spending, Outsized Results
India spends just 0.6% of GDP on research, a fraction of what leaders spend, yet ranks third in scientific papers and sixth in patents. The paradox is real, but the spending gap is too wide to ignore. Here is what the numbers say, and why they matter.
How much do other countries spend on research?
If you lined up every rupee the economy produced in 2020, India put barely six paise of it into research and development. That 0.6% of GDP is the country's R&D intensity, and it is among the lowest for any major economy. Israel, by contrast, set aside 6.3% of its GDP in 2023. South Korea poured in 4.9%. The United States, Japan, and Germany all spent above 3%. Even the world average sat at 2.6%, more than four times India's share. This is not about gross rupees. It is about how much a country prioritises research inside its overall economy. India's share has been stuck in the 0.6–0.7% band since at least 1996. That means for nearly a generation, the fraction of national output going toward creating new knowledge has barely budged, while other countries have pulled far ahead.
The world spends very differently on research
World Bank / UNESCO · gross R&D spending as a share of GDP, 1996 to latest
Israel · 2023 · latest point
India has spent about 0.6% of GDP on research since the 1990s, while Israel, South Korea and China climbed to several times that.
Each line is a country's R&D spending as a share of GDP over time. India's runs almost flat along the bottom near 0.6%. Israel and South Korea sit far above, China rises steeply from a starting point close to India's, and even the world average pulls away to 2.6%.
Why does research spending matter at all?
Spending money on research is not charity. It is one of the highest-return investments a country can make, but with a catch. A firm that invents a new drug or a better chip earns a healthy profit; economists put that private return at roughly 18% a year. But the benefit to society is far larger: rivals learn from the breakthrough, workers grow more productive, and follow-on innovations ripple outward. The measured social return to all R&D sits around 40%, and for agricultural research nearer 50%. Because no single company can capture all of that wider gain, each one spends less on research than the country as a whole needs, which is the basic reason governments everywhere, including India, fund it. The exact size of the payoff is fiercely debated; the direction is not.
Research pays: the question is who reaps it
Economics literature · midpoints of wide ranges, shown to make a point, not as precise measurements
The social return to R&D (around 40% a year) far exceeds the private return a firm captures (about 18%), which is why government funding is critical.
The three bars represent midpoints of economic estimates of annual returns to research. Agricultural R&D sits at about 50%, the highest of the three. The broad social return to all R&D is around 40%. A single firm's private return is much lower, roughly 18%. Because a company cannot capture the full benefit—the knowledge spills over to competitors and the public—it will naturally underinvest. So governments must step in to fund the research that benefits everyone, not just the one who pays for it. These are not precise measurements; they span wide ranges in the academic literature. But the pattern is consistent and underpins the case for India to spend more.
What does the research actually say?
Step back from India for a moment. Does spending on research actually pay off, or is it just a story technocrats tell? On the big question, economists are about as close to agreement as they ever get; the 2025 economics Nobel went to Joel Mokyr, Philippe Aghion and Peter Howitt precisely for explaining how innovation drives growth, Mokyr for the historical preconditions of sustained progress, Aghion and Howitt for the mathematics of "creative destruction", the churn by which better products and methods constantly replace the old. The foundations are older still. Robert Solow showed in 1957 that most of the long-run rise in living standards cannot be explained by simply adding more workers and machines; the unexplained remainder, what economists call total factor productivityTotal factor productivityThe part of economic growth left over once you account for more workers and more machines, often called the Solow residual. Economists read it as a rough measure of technology and know-how. Most of the long-run rise in living standards comes from this, not from sheer accumulation of capital.It is why economists see knowledge and research, not just investment, as the ultimate source of prosperity., is essentially knowledge and technology. Paul Romer's endogenous growthEndogenous growthA family of economic models, associated with Paul Romer, in which long-run growth comes from inside the economy, from deliberate investment in ideas and research, rather than arriving from outside like good weather. It put R&D at the centre of why some countries grow faster than others.It is the theoretical backbone for treating research spending as an engine of growth rather than a luxury. theory, set out in 1990, made the mechanism explicit and won him a Nobel: ideas are unlike ordinary goods because they are non-rivalNon-rivalA good is non-rival if one person using it does not stop anyone else from using it too. A loaf of bread is rival: if you eat it, no one else can. An idea, a formula or a piece of code is non-rival: we can all use it at once. Economists treat knowledge as special for exactly this reason.It is the core reason research compounds and drives long-run growth, and why private firms under-supply it.. Once something is discovered, everyone can use it at once, so research compounds in a way that factories never do.
That same feature is why governments fund research at all. Back in 1962, Kenneth Arrow set out the logic the rest of this article rests on: because a company cannot capture all the gains from what it discovers, rivals copy it, customers benefit, and follow-on inventions spin off, it will always invest less than the country as a whole would want. That gap between what is good for the firm and what is good for society is the textbook case for public research money.
How big is the payoff? The honest answer is large, but hard to pin down. The most-cited survey of the evidence, by Hall, Mairesse and Mohnen, finds private returns to R&D comfortably above those on ordinary investment, and social returns higher still, though variable and imprecisely measured. Bloom, Schankerman and Van Reenen put the social return at roughly twice the private one. Jones and Williams calculated that, even on conservative numbers, a country like the United States should be spending at least four times what it does. The direction is not seriously disputed; the exact multiples are.
Where economists genuinely part ways is on the role of the state. Mariana Mazzucato's The Entrepreneurial State argues that government, not Silicon Valley, seeded the internet, GPS and the touchscreen, and deserves the credit. Critics like Deirdre McCloskey and Alberto Mingardi push back hard, arguing she overstates the state and underplays the private grind of turning inventions into products. Decades of study on whether public R&D crowds private money in or out have reached a verdict that is, honestly, ambivalent. And a newer worry cuts across the whole debate: Nicholas Bloom and colleagues find that ideas are getting harder to find; it now takes more than eighteen times as many researchers to sustain the old pace of computer-chip progress as it did in the early 1970s. For pessimists like Robert Gordon, that is a sign the great wave of innovation is fading; for optimists, it is an argument to put even more people into the lab.
India's own central bank has run the numbers. In its 2024-25 Annual Report, the Reserve Bank of India estimated that a one percentage-point rise in R&D spending lifts India's productivity growth by 0.21 to 0.26 points, and flagged a hopeful wrinkle in the global evidence: middle-income countries like India tend to get more out of domestic research than either poor or rich economies do. The theory, in other words, appears to hold here too.
So what does this mean for India? Mostly that the case for spending more is solid but not a blank cheque. Research pays, and it pays society more than it pays the investor, which is why almost every government funds it. But the returns depend on spending the money well, on having enough researchers to absorb it, and on the private sector eventually carrying its share, the three things the rest of this article shows India is short of.
India's private sector never took over
DST / NSTMIS · share of gross R&D performed by the government bloc vs business · 1990-91 to 2020-21
Government bloc · 2020-21 · latest point
India's business share of R&D climbed through the 2000s to a peak near 45% in 2012-13, then slipped back below 41%, never crossing half.
Two mirrored lines that always sum to 100: the government bloc versus the business sector. Business rises from under a quarter in 2000-01 to about 45% by 2012-13, then falls back. The government bloc, central and state agencies plus universities, has performed the majority of India's research for the entire period.
Has India always spent so little on research?
Not exactly, but the trend line is sobering. India's R&D intensity did nudge upward around 2008–2011, touching roughly 0.8% of GDP, a brief climb that showed what was possible. Then it slid back. By 2020, the figure stood at 0.6%, right where it had been in the mid-1990s. Meanwhile, the world average was no longer the 2% of the 1990s; it had climbed to 2.6%. The dashed line on the chart marks India's own target, set in the 2013 Science, Technology and Innovation PolicyScience, Technology and Innovation Policy (STI) 2013India's policy document that set the goal of raising R&D expenditure to 2% of GDP. It is a declaration of intent, not a binding budget.The 2% target line on several charts traces this unmet aspiration, framing the failure to close the gap.: 2% of GDP. Over a decade later, the country is not closer to that goal. It is further away. This is a policy target, not a measurement, it has never been met. The picture is of a country whose research effort not only stayed flat but slipped relative to a world that was accelerating.
India's research spending peaked over a decade ago
World Bank / UNESCO (India, World) · against the 2% of GDP target set in the 2013 STI Policy
India · 2020 · latest point
India's R&D intensity briefly touched 0.8% around 2008-2011 but slid back to 0.6% by 2020, while the world average climbed to 2.6%.
Two solid lines and one dashed line trace this story. India's R&D spending as a share of GDP rose slowly from 0.6% in 1996, peaked near 0.8% around 2008-2011, then drifted down to 0.6% in 2020. The world average started at 2% and rose to 2.6%. The dashed line at 2% is India's own 2013 policy target—never reached. The gap between India's line and the target has widened, not narrowed, over the past decade. This chart makes clear that India didn't just stay flat in absolute terms; it fell behind a world that was investing more heavily in research.
If the rupee budget is rising, why hasn't the GDP share moved?
Because the economy grew just as fast. In 1995–96, India's gross expenditure on research and development was about ₹7,500 crore. By 2020–21, that had ballooned to roughly ₹1.3 lakh crore, a more than fifteen-fold increase in current rupees. It feels like a surge in ambition. But GDP also multiplied in that same period, and part of the rupee rise was simply inflation eating into the numbers. When you divide a growing numerator by a growing denominator, the share can stand perfectly still. That is exactly what happened: the rupee budget for research kept climbing, but the slice of the national pie it claimed did not. Think of a salaried worker getting a 10% raise in a year when prices rise 10%, the number on the payslip is higher, but what it buys is unchanged. India's research spending has been running to stay in the same place.
More rupees, the same thin sliver
DST / NSTMIS ·
2020-21 · latest point
India's gross R&D expenditure grew from ₹7,500 crore in 1995-96 to ₹1.3 lakh crore in 2020-21, but as a share of GDP it stayed flat because the economy grew just as fast.
The line tracks the total rupees spent on research in India each year, in current prices. It's a steep upward slope: from about ₹7,484 crore in 1995-96 to ₹1,27,381 crore in 2020-21. That looks like a massive increase in ambition. But the denominator—the size of the Indian economy—also ballooned over the same period, and part of the rise is simply inflation. So the share of GDP devoted to R&D barely moved. This chart illustrates the rupee paradox: bigger cheques don't automatically mean a bigger commitment when the whole economy is sprinting ahead.
Who actually does the research in India?
In most rich countries, business does the lion's share of research. Not in India. As of 2020–21, the central government was the single largest performer of R&D, carrying out 43.7% of the total. Private-sector industry came next at 36.4%, with higher education managing only 8.8%, state governments 6.7%, and public-sector industry another 4.4%. Add it up, and the public bloc, centre, states, universities, and government-owned firms, accounts for nearly 60% of all R&D performed in the country. This is the reverse of advanced economies, where companies dominate the lab. It means that India's research agenda is overwhelmingly set by public priorities, not by market demand or industrial competition. That is not automatically bad, but it is structurally different from the model that has driven rapid innovation in South Korea, China, or the United States.
One reason the universities barely feature is structural. India has more than 1,000 universities, but the vast majority are state-run institutions where research has never been the core mission. A 2024 NITI Aayog review of R&D in state universities points to a familiar set of reasons: chronic underfunding, faculty stretched across heavy teaching and administrative loads, long delays in releasing research money, weak support for writing grant proposals, and incentives that reward the quantity of papers over their quality. The report recommends universities ring-fence 5 to 7% of their budgets for research, a measure of how little currently reaches the lab.
In India, the government still does the research
DST / NSTMIS · who performs India's R&D · 2020-21
The central government performs 43.7% of India's R&D, while private industry does only 36.4%, and higher education just 8.8%.
This bar chart breaks down who actually carries out research in India. The largest single performer is the central government, at 43.7% of the national R&D spend. Private-sector industry follows at 36.4%. Higher education institutions account for a thin 8.8%, state governments 6.7%, and public-sector industry 4.4%. If you add up the public sector—centre, states, universities, and government companies—it comes to nearly 60%. In most advanced economies, the private sector is the dominant performer. India's pattern is the reverse, meaning research priorities are shaped more by government missions than by market demand.
Has the private sector been catching up?
Slowly, and then not at all. Through the 2000s India's private sector did take on a growing share of the research, its slice of total R&D climbing from under a quarter in 2000-01 to a peak of about 45% in 2012-13. For a moment it looked like the country might converge on the global model, where business leads. Then the line turned. By 2019-20 the business share had slipped back to 38%, before recovering to 41% in 2020-21. Two decades of trying to pull private money into research have left it stuck below half. (That 41% counts public-sector firms alongside private ones, which is why it runs a little above the 36% private-only share quoted earlier.) The government bloc, central and state agencies plus the universities, still performs the majority of India's research, exactly the opposite of the structure in the economies pulling ahead.
Where India's public research money goes
DST / NSTMIS · central-government R&D by agency · 2020-21
Defence, space, and atomic energy consume about 60% of central-government R&D, leaving civilian and health research thin.
This chart shows how central-government research money is divided among agencies. Defence (DRDO) tops the list at 30.7%. The Department of Space takes 18.4%, the Department of Atomic Energy 11.4%, and agricultural research (ICAR) 12.4%. CSIR accounts for 8.2%. Together, defence, space, and atomic energy soak up roughly 60% of the centre's R&D budget. The Department of Science & Technology gets 6.8%, Biotechnology 4.4%, and medical research (ICMR) only 3.1%. Twelve major agencies cover 84% of all central R&D. This concentration reflects strategic priorities, but it means that areas like health, environment, and university research get leftover crumbs.
Who funds research in other countries?
The single clearest diagnosis of India's research gap sits in a bar chart. In South Korea, 79% of R&D comes from the business sector. In China, 77%. In the United States, 75%. In Germany, 67%. In India, the figure is just 36%. That is not a typo. India's private sector contributes barely a third of research spending, while the rest is borne by government. To be fair, India does not report its data to the OECD's internationally comparable business R&D database, so this 36% figure comes from the Department of Science and Technology's own survey and measures the performer share, not exactly the funder split. But even with that caveat, the order-of-magnitude difference is unmistakable. When the engine of innovation in the rest of the world is private enterprise, India's research landscape is still largely a government project.
Everywhere else, business pays for research
OECD / WIPO / NITI Aayog · business-enterprise share of R&D · India is its private-industry share (DST 2020-21)
Business funds 79% of R&D in South Korea, 77% in China, 75% in the US, and 67% in Germany, but only 36% in India.
Five bars lay out the share of research that comes from the business sector. South Korea leads at 79%, followed by China (77%), the United States (75%), and Germany (67%). India's bar is dramatically shorter, at just 36%. This is the clearest single structural gap between India and innovation-led economies. In every country that has transformed its economy through R&D, companies have been the main engine. India, instead, remains heavily reliant on government spending. The caveat: India's 36% is from the DST's performer survey, not the OECD's strictly comparable business-enterprise R&D definition, but the order-of-magnitude difference is unmistakable.
What does the government spend its research money on?
When you look inside the central government's R&D budget, a clear pattern emerges: the money is concentrated among a handful of agencies, and it tilts heavily toward strategic missions. In 2020–21, the Defence Research and Development Organisation (DRDO) alone consumed 30.7% of central R&D spending. The Department of Space took 18.4%, the Department of Atomic Energy 11.4%, and the Council of Scientific and Industrial Research (CSIR) another 8.2%. Add defence, space, and atomic energy together, and they commanded about 60% of the central research purse. The Indian Council of Agricultural Research (ICAR) got 12.4%, reflecting agriculture's importance, but other civilian agencies, the Department of Science and Technology, the Department of Biotechnology, the Indian Council of Medical Research, received much smaller shares. In all, twelve major agencies accounted for 84% of the centre's R&D. This is a deliberate national choice, reflecting security and prestige goals. But it leaves health, university, and civilian industrial research remarkably thin.
And where the private money goes
DST / NSTMIS · leading industry groups by share of industrial R&D · 2020-21
One sector, drugs and pharmaceuticals, is about a third of all of India's industrial R&D; everything else trails far behind.
Bars show each industry's share of industrial R&D in 2020-21. Drugs and pharmaceuticals leads at 33.6%, then information technology (9.9%), transportation (7.7%), defence industries (7.3%) and biotechnology (4%). India's private research effort is concentrated in a few globally competitive industries.
And what does the private sector research?
The private research that does happen is strikingly narrow. In 2020-21 a single sector, drugs and pharmaceuticals, made up about a third (33.6%) of all industrial R&D. Information technology was a distant second at 9.9%, followed by transportation at 7.7%, defence industries at 7.3%, and biotechnology at 4%. India's private research effort leans heavily on one globally competitive industry, generic pharma, with a long tail behind it. That concentration is a strength and a fragility at once: it reflects where Indian firms have found a real edge, but it also means whole swathes of the economy do almost no research of their own.
And yet: third in the world for research papers
NSF Science & Engineering Indicators / Scopus · scientific & technical publications · 2022
India published 278,000 scientific papers in 2022, ranking third behind China (1 million) and the US (721,000).
On this bar chart, publication volume pops. China leads with about one million papers in 2022. The United States published around 721,000. India is third at about 278,000, having overtaken the United Kingdom, Germany, and Japan. This is the efficiency paradox: a country that spends so little on R&D manages to pump out a lot of visible output. However, this is a count of papers, not a measure of their influence. Citation-based indices typically rank India lower. And a significant portion of this output stems from research infrastructure and training built years ago. Still, the absolute number proves that India's research system is not dormant—it's just starved.
How did China and South Korea get so far ahead?
In the mid-1990s, India and China were not far apart. Both spent around 0.6% of GDP on research. South Korea was already higher at 2.1%, but the gap was measurable, not astronomical. Then their paths split. By 2023, China had raised its R&D intensity to 2.6%, more than four times its starting point and nearly exactly the world average. South Korea had shot up to 4.9%, approaching Israel's territory. India, after a brief flirtation with 0.8% over a decade ago, drifted back to 0.6%. This is not a story of one bad year; it is two generations of deliberately different choices. The multi-line chart makes it plain: two lines climb steeply, one line stays flat, and the world average leaves it behind. Because China's economy is also much larger than India's, the absolute spending gap is even wider, but that is a story for another chart.
China and Korea pulled away
World Bank / UNESCO · R&D spending as a share of GDP, India vs the risers
South Korea · 2023 · latest point
While India's R&D intensity stayed near 0.6%, China rose from 0.6% to 2.6%, and Korea from 2.1% to 4.9% over the same period.
Four lines on one chart tell a story of diverging ambitions. India starts at 0.6% of GDP in 1996 and ends at 0.6% in 2020, with a small peak in between. China starts right next to India at 0.6% and climbs steadily to 2.6% by 2023. South Korea is already higher at 2.1% in 1996 but rockets to 4.9%. The world average moves from 2% to 2.6%. India's line flatlines while everyone else's rises. That two-generation split shows how deliberate national choices have reshaped the global innovation landscape. India's stagnation is not a fixed trait; it's a policy choice that other countries, once in similar positions, did not make.
What about countries more like India?
The giants are easy to wave away: of course India cannot match a superpower. So look instead at the economies that started where India did, or behind it. In the mid-1990s India devoted a larger share of its GDP to research than Malaysia, Thailand, Vietnam, or Indonesia, all of which sat below a quarter of a percent. Then they moved, and India did not. By the early 2020s Malaysia had reached about 1.0% of GDP and Thailand about 0.9%, both comfortably past India's 0.6%. Vietnam roughly doubled its share, halving the distance to India. Only Indonesia and Pakistan still sit clearly below. The uncomfortable point is that India's flat line is not just a failure to keep pace with China and Korea; it is a country that has been overtaken by smaller economies that decided research was worth paying for.
Even Southeast Asia is pulling ahead
World Bank / UNESCO · gross R&D spending as a share of GDP · India vs its Asian neighbours
India · 2020 · latest point
India spent a bigger share of GDP on research than Malaysia, Thailand, Vietnam and Indonesia in the 1990s. By the 2020s Malaysia (about 1.0%) and Thailand (about 0.9%) had overtaken it, while India stayed near 0.6%.
Six lines, all starting near or below India in the mid-1990s. Malaysia climbs from about 0.2% to 1.0% of GDP and Thailand from 0.1% to 0.9%, both passing India. Vietnam roughly doubles to 0.4%, halving its gap with India. Indonesia and Pakistan rise a little but stay below. India's line barely moves the whole time. The point is not that India trails superpowers; it is that economies it once out-spent have left it behind.
How the countries that pulled ahead did it
None of these countries got rich first and then bought research; most spent their way up while still developing. The common thread is a choice India has not made: to treat research as a sustained national priority, and to pull private money in behind the state.
South Korea is the clearest case. In the mid-1990s it spent about 2% of GDP on research; today it spends close to 5%, second only to Israel. It did this by setting hard national targets and leaning on its big industrial groups, Samsung, Hyundai and LG, to perform most of the actual research, with government co-funding and a relentless push from cheap manufacturing toward chips, cars and electronics. China started exactly where India did, at 0.6% in 1996, and chose the opposite of standing still: a state-directed climb to 2.6%, tied to its manufacturing base and backed by talent-recruitment programmes and sheer scale. Israel, the world's most research-intensive economy at over 6% of GDP, took a third route, leaning on defence research spillovers, a wave of scientist immigration and the densest venture-capital ecosystem on earth: intensity rather than size.
The other two are honest about the limits. Singapore built a research base almost from nothing through deliberate state investment and its A*STAR agency, reaching 2.6% of GDP by 2008, then drifted back toward 1.8% as its economy grew faster than its labs. Finland rode an education-first, publicly-funded ramp to nearly 3.7% around 2009, anchored by the Nokia cluster, then slid below 3.2% when Nokia faltered. Both still run research economies several times more intensive than India's, but they are a reminder that research leadership is not permanent; it has to be renewed.
The pattern across all five is still unmistakable. Each made a deliberate, decades-long bet on research that India, so far, has not, and even the ones that have slipped did so from heights India has never reached.
Everyone made the climb except India
World Bank / UNESCO · R&D spending as a share of GDP since 1996 · one panel per country
Each chart is one city, 1996–2023, auto-scaled to its own range so the trend is visible — read the slope, not the height. The big number is the 2023 value; the small coloured number is the change since the 1996s (red = warmer, blue = cooler). Colour of the line shows how high the 2023 value sits across all cities.
Five economies that began at or below today's India, or alongside it, all built far more research-intensive economies; India's line stayed flat.
Each small chart is one country's R&D spending as a share of GDP since 1996. Israel and South Korea climbed to the top of the world; China rose from India's exact 1996 starting point to 2.6%; Singapore and Finland rose then eased back, but both remain several times above India. India's own panel barely moves off 0.6%.
How many researchers does India have?
Money is only one half of the research equation. The other is people. And here, India's numbers are genuinely startling. According to UNESCO, India had about 259 researchers (full-time equivalent) per million people in 2020. South Korea had 9,472. Germany had 5,926, Japan 5,609, and the United States nearly 4,937. Even China, with its enormous population, managed 2,107. The world average was 1,651. Starker still, India sat below Vietnam (836) and South Africa (444). The Department of Science and Technology's own series, using a slightly different definition, puts the figure at 262 for 2020. The broad message is the same in both: for every million Indians, there are barely 260 scientists and engineers engaged in creating new knowledge. For a country that wants to be a knowledge superpower, the human foundation is alarmingly thin. In absolute numbers, about 5.55 lakh people worked in India's R&D establishments as of April 2021, of whom roughly 3.6 lakh were directly engaged in research; the rest were technical, administrative and support staff.
India has strikingly few researchers
World Bank / UNESCO · full-time-equivalent researchers per million population · latest available per country
India has only 259 researchers per million people, versus 9,472 in South Korea and even 836 in Vietnam.
This bar chart compares the density of researchers—scientists and engineers engaged in R&D—across countries. South Korea is off the charts at 9,472 per million. Germany (5,926), Japan (5,609), and the United States (4,937) all have several thousand. China, despite its huge population, scores 2,107. The world average is 1,651. Then the bars shrink: Vietnam 836, South Africa 444, and India only 259. That's fewer than half of South Africa's and a fraction of Vietnam's. For every million Indians, barely 260 are working full-time to create new knowledge. This people gap is just as serious as the spending gap; research needs both rupees and brains.
Is India's researcher base growing?
Yes, it is, but the starting point was so low that the climb barely registers globally. India's own official series, drawn from DST surveys, shows the number of researchers per million population rising from 110 in 2000, to 218 in 2015, to 255 in 2017, and to 262 in 2020. Measured from the turn of the century, the density more than doubled. That is real progress, and it reflects the expansion of higher education and PhD output over those decades. But doubling a very small number still leaves it very small. At this rate, India would need decades to catch up to where China is today, even assuming China stops growing its own researcher pool, which, of course, it has not. The trend is positive, but the absolute stock remains a deep structural weakness, not a temporary blip. The pipeline is widening at the top, too: India now awards the third-most science and engineering PhDs in the world, around 24,500 a year, behind only the United States and China. But a large doctoral output sitting on a tiny per-capita researcher base shows how thinly that talent spreads once it enters the workforce.
India is adding researchers, but from a very low base
DST / NSTMIS ·
2020 · latest point
India's researcher density has more than doubled from 110 per million in 2000 to 262 in 2020, but remains negligible by global standards.
This line chart zooms in on India's own domestic series. It shows the number of researchers per million people at a few selected points: 110 in 2000, 218 in 2015, 255 in 2017, and 262 in 2020. The upward movement is genuine and reflects the expansion of higher education and PhD output. The density more than doubled in twenty years. But the absolute numbers are still tiny. At this pace, catching up to today's world average would take many decades, assuming other countries stop growing—which they won't. Progress is real, but the base was so low that even a doubling doesn't change India's global rank.
And who are the researchers?
Overwhelmingly men. Of the roughly 3.6 lakh people directly engaged in R&D in India as of April 2021, only 18.6% were women. The picture is a little better among funded project leads, where women's share of extramural researchExtramural researchResearch funded by one body but carried out by another, typically a government science agency giving grants to researchers in universities and institutes outside its own labs ('extramural' means outside the walls). It is the main channel through which public money reaches academic research in India.The gender and grant 'principal investigator' figures come from these extramural schemes. grants roughly doubled from 13% in 2000-01 to about a quarter by the late 2000s. But it then stalled: after touching 33% in 2013-14, the share drifted back to 25% by 2019-20. In absolute terms that is real progress, 848 women led grants in 2019-20 against just 232 two decades earlier, but a research workforce that is three-quarters male, and no longer closing the gap, leaves a lot of talent unused.
And the researchers are mostly men
DST / NSTMIS ·
2019-20 · latest point
Women lead only about a quarter of funded research projects, and that share peaked in 2013-14 and has since slipped.
The line is women's share of principal investigators on extramural grants. It jumps from 13% in 2000-01 into the mid-20s, peaks at 33% in 2013-14, then drifts back to 25% by 2019-20. Across all R&D personnel, women are 18.6%.
What does the spending gap look like in absolute money terms?
Percentages can be misleading. India's 0.6% of GDP seems like a modest difference from, say, China's 2.6%, a factor of about four. But GDP sizes are not equal. When you convert R&D spending into purchasing-power-parity dollars, a common currency that adjusts for what a dollar can buy in each country, the gap explodes. In 2020–21, India's total R&D spend was about 58 billion PPP dollars. The United States spent about 760 billion, and China about 860 billion, both in 2024. That means China, with four times India's R&D share of GDP, spent nearly fifteen times as much in absolute terms. The United States outspent India by a factor of thirteen. This is not just a question of intensity; it is a question of how much actual brainpower, equipment, and experimentation money can buy. GDP shares understate the distance between India and the research superpowers. In absolute scale, India is not even in the same weight class.
The absolute scale of the gap
OECD MSTI (China, US · 2024) and DST / UNESCO (India · 2020-21) · purchasing-power-parity dollars
India's R&D spend of $58 billion (PPP, 2020-21) is dwarfed by China's $860 billion and the US's $760 billion.
Three bars compare absolute R&D spending in purchasing-power-parity dollars, which adjust for price differences. China, in 2024, spent about $860 billion. The United States spent $760 billion. India, in 2020-21, spent roughly $58 billion. That means China spends nearly fifteen times as much on research as India, and the US spends thirteen times as much. The GDP-share comparison (2.6% vs 0.6%) suggested a four-to-one difference; in absolute money the gap is far wider because China's economy is also about four times as large as India's. This bar chart makes it visual: India's bar is barely visible next to the others. Scale matters: more money means more labs, more equipment, more researchers, and more shots on goal.
But doesn't India produce a lot of research anyway?
This is the paradox that keeps the conversation from being a simple story of neglect. In 2022, India published about 278,000 scientific papers. That placed it third in the world, behind only China (about one million) and the United States (about 721,000). India has overtaken the United Kingdom, Germany, and Japan in raw publication count. On the face of it, a country spending like an also-ran is producing like a heavyweight. But the honest caveats matter. Publication counts measure volume, not influence. On citation-based rankings that track how often other researchers build on your work, India scores lower. More importantly, today's paper output partly reflects research capacity and people trained years ago, not the current spending level. A country can coast on past investments for a while. The question is whether it can sustain that when the pipeline of new investment is so thin. By one widely used database, the US National Science Foundation's, India's share of the world's scientific papers has roughly doubled, from 3.1% in 2010 to 5.1% in 2020, growing more than twice as fast as the world average.
The research money that doesn't get spent
Open Budgets India · share of the budgeted R&D allocation actually used
DST used only 82% of its budget on average, but its R&D head spent just 34% in 2023-24, and ANRF used only 13% of its allocation.
This chart shows the share of budgeted allocation actually spent. Over 2018-24, the Department of Science and Technology managed to spend an average of 82% of its total budget. That might sound okay, but its specific R&D head in 2023-24 spent only 34%—barely a third. The new Anusandhan National Research Foundation got off to a dismal start, using just 13% of its budget in its first year. The reasons are chronic: rigid government financial rules that claw back unspent money at year-end, long procurement delays, GST on research equipment, and a general administrative drag. This is not just a funding problem; it's an execution problem. Even the limited money that is allocated often fails to reach labs and scientists.
What about patents?
If publications are the measure of knowledge creation, patents are the closest we get to measuring invention aimed at commercial use. And here, the trend is genuinely bright. In 2013, Indian residents filed only 10,669 patent applications at the Indian Patent Office, while foreign applicants filed 32,362. Over the next decade, resident filings climbed sharply, non-resident ones crept up slowly. In 2022, for the first time, Indians filed more patent applications in their own country than foreigners did. By 2024, resident filings reached 63,217, about 60% of all applications, while non-resident filings stood at 41,940. India has also risen to sixth place globally for total patent filings. This is a genuine shift from dependency toward domestic inventive ambition. But a patent application is still a piece of paper. It is an input, not proof that an invention has been commercialised or that it will create value. The valley between filing and a product on the shelf remains wide.
Indians now file most of their own patents
WIPO / World Bank · applications filed at the Indian Patent Office, by applicant
Resident (Indian) applicants · 2024 · latest point
Resident patent filings in India surged from 10,669 in 2013 to 63,217 in 2024, overtaking non-resident filings for the first time in 2022.
Two lines chart a remarkable shift. In 2013, foreign applicants dominated, filing 32,362 patents compared to just 10,669 from Indian residents. Over the next decade, resident filings grew rapidly, crossing non-resident in 2022. By 2024, Indians filed 63,217 applications, while foreigners filed 41,940. That means domestic inventors now account for about 60% of all patent activity at the Indian Patent Office. India is now sixth in the world for total patent filings. This is a genuine bright spot, signaling rising domestic inventive ambition. But a patent filing is not a granted patent, and a granted patent is not a commercial product. The next step—turning filings into marketable innovations—remains the weak link.
Is India more efficient at turning money into innovation?
The numbers suggest yes, and they have been doing so for years. Every year, the World Intellectual Property Organization's Global InnovationGlobal Innovation Index (GII)A composite ranking by WIPO that scores countries on innovation inputs (institutions, human capital, infrastructure, market and business sophistication) and outputs (knowledge and technology outputs, creative outputs). Overperformance means outputs rank higher than inputs.India's consistent overperformance is the clearest evidence of R&D efficiency, but it does not justify underinvestment. Index (GII) ranks roughly 140 economies on both innovation inputs (things like institutions, human capital, infrastructure, and R&D spending) and innovation outputs (knowledge creation, technology, and creative goods). In the 2025 edition, India ranked 52nd on inputs and 32nd on outputs. That gap, outputs performing twenty places higher than inputs, makes India a consistent "overperformer," a label it has held for about fifteen years. It means that relative to the money and infrastructure it puts in, India produces unexpectedly strong results. This is not a reason to celebrate low spending, but it is a powerful argument that the country's research system is not wasteful. Efficiency means that if more money were available, it would likely be put to good use. The existing talent and institutions could absorb far more than they currently receive. There is a flip side worth naming. Because India has so few researchers, the money it does spend, divided across them, is not unusually low: its R&D spend per researcher, about 160,000 PPP dollars in 2020-21, actually runs ahead of Russia, Hungary and Mexico. India's problem is less that each researcher is starved of funds than that there are far too few researchers, and far too little private money, to begin with.
India spends like 50th, innovates like 30th
WIPO Global Innovation Index 2025 · India's rank on innovation inputs vs outputs
India ranks 52nd on innovation inputs but 32nd on innovation outputs, a gap that has held for over a decade.
Two bars show India's rank out of 139 economies in the Global Innovation Index 2025. On innovation inputs—which include R&D spending, institutions, and human capital—India sits at 52. On innovation outputs—knowledge creation, technology, and creative goods—it ranks 32. That twenty-place gap makes India a classic 'overperformer'. It consistently produces more innovation than its resource base would predict. This pattern has been stable for about fifteen years. The efficiency is real, but it does not make the input deficit unimportant. It suggests that if India did invest more, the system could absorb and convert that money into results. Overperformance is not a substitute for adequate investment.
What are India's best examples of research on a budget?
The most visible example is the Indian Space Research Organisation, or ISROISRO (Indian Space Research Organisation)India's national space agency, known for low-cost, high-impact missions. It exemplifies a mission-driven, frugal approach to public R&D.ISRO's cost-effectiveness shows that when research is well-focused, India can achieve world-class results on a shoestring, suggesting that more money could be used well.. When ISRO sent its Mars orbiter, Mangalyaan, to the red planet in 2014, the mission cost about $74 million, roughly the budget of a mid-range Hollywood film. Its Chandrayaan-3, which landed near the Moon's south pole in 2023, cost roughly $75 million. Compare that to NASA's MAVEN Mars orbiter at $582 million or Russia's Luna-25 Moon mission at $133 million, which ended in a crash. The ISRO model is mission-focused, frugal, and autonomous; it does not try to do everything, but what it does, it does for a fraction of the cost. India's vaccine-manufacturing base, led by the Serum Institute and Bharat Biotech, working with the Indian Council of Medical Research on Covaxin, is another public-private win that delivered at scale. CSIR's work on generic-drug process chemistry has saved millions of lives. These are not accidents. They show that when Indian research is well-directed and accountable, it can achieve world-class results on shoestring budgets.
India does science cheap: the ISRO model
Mission budgets · India's space programme vs comparable foreign missions · approximate
India's Mars mission cost $74 million and its Moon landing $75 million, while a comparable NASA Mars mission cost $582 million.
Four bars compare headline mission costs. India's Mangalyaan to Mars: $74 million. Chandrayaan-3 Moon landing: $75 million. Russia's Luna-25, which crashed: $133 million. NASA's MAVEN Mars orbiter: $582 million. The Indian missions cost an order of magnitude less. ISRO's frugal, mission-focused, and autonomous approach squeezes extraordinary value from limited budgets. The same pattern shows up in vaccine manufacturing, where India's Serum Institute and Bharat Biotech delivered at scale, and in CSIR's work on generic drugs. These are not outliers; they demonstrate a genuine national capability to do high-quality research on a shoestring. The headline costs exclude shared infrastructure and prior decades of investment, so the point is the relative frugality, not an exact accounting.
What goes wrong with India's research spending?
For all the frugal successes, the system has chronic leaks. One of the most glaring is money that never gets out the door. Over the five years from 2018–19 to 2023–24, the Department of Science and Technology used, on average, only 82% of its budgeted allocation. In the specific case of its R&D head in 2023–24, utilisation collapsed to just 34%. The new Anusandhan National Research FoundationAnusandhan National Research Foundation (ANRF)A new research funding body created in 2023 to coordinate and fund R&D across disciplines, with a target corpus of ₹50,000 crore over five years, of which only ₹14,000 crore is committed government money. It is meant to attract large private contributions.ANRF is India's main institutional bet to fix the funding gap, but its success hinges on private money that has yet to materialise. (ANRF) managed to spend barely 13% of its allocation in the same year. The reasons are structural: government financial rules that withdraw unspent funds at year-end, long procurement delays, GST on research equipment, and a general culture where spending is hard and saving is easier. Beyond underutilisation, there is the "valley of death" between a lab invention and a market product, where most Indian patents never get commercialised. Long-running prestige projects also weigh on resources: the Kaveri jet engine, in development for roughly 39 years, still does not power a fighter plane. These are documented problems, not proof that the whole system is wasteful. But they show that fixing the research gap is not just about writing bigger budget cheques.
The new bets on Indian research
ANRF / RDI Scheme · the headline new funding commitments
India's new ANRF targets ₹50,000 crore over five years, but only ₹14,000 crore is committed government money; a separate ₹1 lakh crore RDI fund aims to lend to private projects.
Three bars show the headline figures. The Anusandhan National Research Foundation (ANRF) has a government commitment of ₹14,000 crore, but its total target is ₹50,000 crore—the rest, ₹36,000 crore, is supposed to come from private and other sources. The RDI Fund, a separate vehicle, offers a ₹1 lakh crore financing corpus, but it lends or takes equity, rather than giving grants. These are ambitious numbers that signal a desire to fix the gap. But the private share is aspirational, not committed, and past efforts to attract private R&D money have failed. The RDI fund's model also means it will favor projects close to market readiness, not basic science. Whether these bets pay off depends entirely on private money turning up.
What is India doing to fix the gap?
The government's main answer is a set of large new funding vehicles. The Anusandhan National Research Foundation (ANRF), created in 2023, targets a total corpus of ₹50,000 crore over five years, of which only ₹14,000 crore is committed government money; the rest, ₹36,000 crore, is meant to come from industry, philanthropy and other private sources. Alongside it sits the bigger bet: the Research, Development and Innovation (RDI) Scheme, a ₹1 lakh crore fund the Union Cabinet approved in July 2025 to pull private money into deep-tech and strategic sectors. Rather than hand out grants, it works through a special fund parked inside ANRF that lends to private projects as long-term, low-or-no-interest capital; the Budget for 2025-26 set aside the first ₹20,000 crore.
Both are ambitious, and they signal that policymakers understand the gap and are trying new models. But the arithmetic is sobering. Even if ANRF's full corpus were met, the annual addition would be only about ₹10,000 crore, and the RDI fund's first tranche is ₹20,000 crore, together a small fraction of what it would take to push India's R&D intensity from 0.6% to even 1%, let alone 2%. And both bets lean heavily on private money that, so far, has been conspicuously absent: a fund that lends to companies only works if companies actually borrow and spend. Until that money turns up, the 2% target stays aspirational.
How should you read these numbers?
Every statistic on this page comes from specific, named sources, and each carries important caveats. The cross-country R&D intensity and researcher density figures are drawn from the World Bank and UNESCO Institute for Statistics, mostly for 2020–2023. India's detailed rupee numbers, the sector and agency splits, and the domestic researcher count come from the Department of Science and Technology's NSTMIS, specifically the "R&D Statistics at a Glance 2022–23" and the S&T Indicators Tables, whose latest year is 2020–21. That means India's official R&D data lags about five years behind the year you are reading this. NITI Aayog's 2025–26 reports provide diagnosis and reaffirm the targets, but rely on the same base data. The business-sector comparison is approximate: India is absent from the OECD's comparable database, so its 36% figure is a DST performer-share estimate rather than the standard OECD business-enterprise R&D definition. Patent figures are from WIPO's World Intellectual Property Indicators, and publication counts from the NSF Science and Engineering Indicators via Scopus. The returns to R&D are midpoints of wide and contested academic ranges, not observed levels. Multiple factors move together, weak private investment, thin university research, fragmented funding, brain drain, and concentration among a few institutes, so no single factor is proven to be the cause. What the data collectively show is a clear, persistent underinvestment against India's own goals and global peers, side-by-side with a remarkable ability to squeeze output from limited input. The precise figures will shift with new surveys, but the broad picture is robust and has been stable for decades.