by Jason Crawford · April 23, 2020 · 5 min read
In my last post I wrote about the importance of funding models to progress. Here I want to survey the major types of funding models particularly for science and invention. This list is probably incomplete, but it’s a starting point.
There are different ways we could classify funding models, for instance by:
I don’t find any of those axes to be helpful as the primary way to organize the topics in this area, in part because many types of institutions both perform research in-house and fund research externally. It also doesn’t make sense to examine separately every combination of the three attributes above.
To the extent I am able to choose a primary axis, I think the most helpful one is the type of institution involved. What follows is the breakdown I’m working with now.
A remarkable amount of progress in history has been made by individuals researching or inventing outside the context of any formal institution. Models to support this include:
Independent wealth. Often this means family inheritance, as with the many “gentleman scientists” of the 18th and 19th centuries. It can also come from business success: Leo Baekeland invented the first synthetic plastic, “Bakelite”, after he sold his photographic paper company to Kodak.
Patronage. Galileo was supported by the Medici; when he discovered the first moons of Jupiter he named them the Medicean Stars. Denis Papin, who demonstrated a steam piston in the 1600s, was supported by Huygens and Boyle, among others. The Hapsburg emperor Leopold I funded enough scientists that the German Academy is known as the Leopoldina.
Sinecures or other relatively undemanding jobs that leave time to experiment and tinker. Karl von Drais invented an early proto-bicycle while a forestry official. Christopher Sholes, a customs collector, invented a typewriter. William Harvey discovered the circulation of blood while working as a physician to the aristocracy. Antonie van Leeuwenhoek, microscope pioneer, was a chamberlain in Delft.
Those without any of the above means still sometimes managed to make progress through working overtime, creative side hustles, or risking debtor’s jail—I gave several examples in my post on early American inventors.
These models were important up until perhaps the 1800s, when much research and invention was done by individuals. Now that most R&D is done in institutions, models of institutional funding have become relatively more important.
University. When most people think “research”, the default is probably to think of a university. In addition to doing the lion’s share of basic research in the sciences, universities often prototype inventions that are then productionized by other institutions. The University of Karlsruhe gave us the principle behind synthetic fertilizer. Oxford performed the first clinical trials on penicillin. The University of Pennsylvania built ENIAC, the prototype of the computer.
Foundations and other private non-profit organizations. Some of these perform research in-house, such as the Fred Hutchinson Cancer Research Center. Others simply fund and coordinate research at other institutions, such as the National Foundation for Infantile Paralysis, which was dedicated to polio. Some do both, such as the Simons Foundation. Some are not dedicated exclusively to research, but fund it alongside other causes, such as the Gates Foundation.
Military. During World War II, the National Defense Research Council and later the Office of Scientific Research and Development supervised a variety of projects from radar to penicillin to the atomic bomb. The OSRD was dissolved after the war, but a decade later, after the Soviets launched the satellite Sputnik, the US once again decided to prioritize military research. DARPA, created soon after, gave us the first computer networks (the precursor to today’s Internet).
Other government agencies for public benefit. The US National Institutes for Health, with a budget of over $40B, is by far the largest funder of health research in the world. Other US agencies involved in research include the National Science Foundation, NASA, and the Department of Energy. These agencies often do research in-house and also make grants to universities and corporations.
Corporate research labs. The value of research to business was recognized as early as the 1800s, when chemistry began to have applications to many industries, from dyes to steel to cement. In Denmark, the Carlsburg Laboratory was created to study fermentation for brewers. In Germany, Bayer gave us aspirin, and BASF productionized synthetic fertilizer. In the US, Edison’s lab invented the light bulb (yes, they did), Bell Labs gave us the transistor among many other innovations, and Xerox PARC pioneered the personal computer.
For-profit. Companies can also do R&D more directly as a part of product development or operational improvement. I distinguish this from corporate research labs in that the latter are generally distinct entities under a successful, established corporate parent, funded out of the parent’s free cash flow; they have a more long-term focus, are managed less by near-term results, and often publish papers. For-profit businesses naturally tend to focus more on applied research, invention, and product development, but they have been known to contribute even to basic science. Josiah Wedgewood, the ceramics manufacturer, advanced the field of thermometry with a method to measure very high temperatures (as in a kiln); the t-test, a basic statistical technique, was invented by an employee of the Guiness brewery.
The models above are based on the types of institutions that fund research, or that organize it. However, a few special allocation mechanisms deserve consideration as models in their own right:
Prizes. The Longitude Prize was established in the 1700s for a solution to the longitude problem (partially won by John Harrison), the Alhumbert Prize in the 1800s for progress on the question of spontaneous generation (won by Louis Pasteur), and the Orteig Prize in the 1900s for cross-Atlantic flight (won by Charles Lindbergh). Arguably you could include even institutions such as the Nobel Prize, to the extent that the recipient is able to use the prize money to support independent research. The difference between prizes and grants is not only that prizes are given after the accomplishment, but also that they obviate the application and review process: a prize is generally open to all and is granted upon achievement of an objectively defined goal.
Lotteries. The grant process is time-consuming for both applicants and reviewers, and it’s not clear that it results in good decisions or even objective ones. Because of this, some have suggested giving out at least a portion of grant money by random lottery (probably after a quick screening for minimal quality). The Health Research Council of New Zealand has experimented with this, as well as the Swiss National Science Foundation; the results are still being evaluated.
There are other lesser-known models as well, such as advance market commitments or patent buyouts, described here. These are similar to prizes, but with a sales contract instead of a cash reward.
There are also hybrid approaches. For instance, some grants or investments are divided into tranches that are awarded on meeting certain milestones. This combines elements of a basic grant/investment and a prize.
Another way to think about funding models is “push” vs. “pull”. In a “push” model, you aggregate funding and adopt a mechanism to get it to researchers, typically a grant program. Usually you announce the program and state your purpose and goals. In a “pull” model, you create an institution to house research, and then you seek and attract funding. Many institutions combine both, e.g., NIH has both a grant program and in-house research.
Push and pull models work together. A foundation (push) may give a grant to a university (pull). Or a prize (push) may justify investment from a for-profit company (pull).
With this framework, we can start to ask some questions, like:
Social media link image credit: Wikimedia / Graham Norrie
Get posts by email:
Don't show again