In 1810, Wilhelm von Humboldt founded the University of Berlin, and in doing so, invented the modern research university itself. Before that, a university in Europe was mostly a teaching operation; professors recited settled knowledge, students memorized it, and nobody expected new knowledge to come out the other end.
Humboldt's model rested on a simple, radical idea: Einheit von Forschung und Lehre, the unity of research and teaching. A professor's job was not just to teach what was known, it was to produce what was not known yet, with students learning by doing research alongside them rather than simply sitting through lectures about it. Add Lehrfreiheit, the freedom to teach and research without political interference, and Lernfreiheit, a student's freedom to shape their own course of study, and the result is, more or less, the template every serious research university on earth still runs on today.
The world noticed quickly. Johns Hopkins, founded in 1876, was the first American university built largely on the German model (a hybrid of the German research model and the British residential-college tradition, though overwhelmingly German in spirit); its founding president, Daniel Coit Gilman, had studied the German system directly and imported it, seminar-style teaching and research as the actual point of the institution, not an afterthought to it. Harvard, Cornell, and Duke reshaped themselves around the same idea soon after.
(There is an earlier piece by Shreyasi Singh that covers a bit of that same import: worth a read.)
For a few decades, the bet paid off spectacularly. In the decades before Hitler barred Germans from accepting the prize, German scientists took home close to a quarter of all Nobel prizes awarded, a dominance Nature itself was still tallying admiringly into the 1930s.
And then Germany did something instructive: it destroyed its own advantage within a single generation. Nazi-era purges and boycotts drove out Jewish scientists by the hundreds, many of whom went on to rebuild American science instead. The lesson is not simply "build the right institution." It is that the institution is not self-sustaining; it survives only as long as the surrounding political culture allows it to.
The more interesting move, though, came after the war. Germany did not just rebuild Humboldt's university. It split the model in two.
The Max Planck Society, founded in 1948 as successor to the older Kaiser Wilhelm Society, does pure research, and pure research only, deliberately kept outside the university system so its scientists are not burdened with teaching loads. It runs 85 institutes today, on a €2.15 billion annual budget, and counts 31 Nobel laureates in its own ranks. It exists to answer one question: what is true, with no obligation to make it useful to anyone.
The Fraunhofer Society, founded a year later in 1949, is the deliberate answer to the other question: what happens to an idea after the paper is published. It is Germany's applied-research counterpart, contract-funded by industry rather than by the state. It runs 75 institutes, with €3.6 billion in business volume in 2025, more than seventy percent of it earned rather than granted. A company with a specific engineering problem pays Fraunhofer to solve it. That is the entire model.
Germany, in other words, did not just build one great institution. It eventually recognised that a pure-research culture and an applied, industry-facing one need two different institutions with two different incentive structures, and it built both, deliberately, a generation apart.
India has made neither move at real scale. India's total R&D spend sits at 0.84 percent of GDP as of FY24, up from 0.66 percent a few years earlier, but still a fraction of what a country like Germany spends.
India also lacks a Fraunhofer-equivalent: a nationally scaled, industry-contract-funded applied research network sitting between pure academic research and private industry R&D. The IITs' own tech-transfer efforts, such as IIT Delhi's FITT, are real but small, single-institution versions of exactly this idea, not a national system.
CSIR is India's closest institutional analogue to a Fraunhofer, roughly three dozen government labs meant to conduct applied, industry-relevant research. But in 2015 the government directed CSIR to self-finance 30 to 50 percent of its own budget by commercialising its own technology, and by 2017-18, after salaries and pensions were paid out of a ₹4,063 crore allocation, only ₹202 crore remained for laboratories and new research.
That is not Fraunhofer's model. That is a government lab being quietly starved while being asked to behave like a business it was never built to be.
The IITs are not formally CSIR labs, but they are pushed to behave like one regardless; government funding schemes and consultancy mandates increasingly expect them to run applied, industry-facing research on top of teaching, without ever being given a separate institution, or a separate funding logic, built for that job. India's IITs are consequently trying to be three things at once: a Humboldt-style teaching-and-research university, a Fraunhofer-style commercialisation engine, and something like a CSIR-style national lab, without ever building the second and third pieces at the scale Germany did. That is not a resourcing problem solvable by giving the IITs more money to do more things. It is a design problem; the same institution cannot easily hold all three incentive structures at once, and Germany's own history makes that case.
Germany already ran this experiment and has the answer. Pure research and applied research are not the same job, and asking one institution to do both, as India currently asks its best universities to, under-serves all three missions equally.
What India has not yet built is not another IIT. It is an actual Fraunhofer: funded by industry contracts rather than government grace, with the freedom to fail that comes from not also needing to teach undergraduates or fill a bureaucrat's five-year plan.
