How a Sovereign Wealth Fund Could Help Jump-Start Science Policy
We Should Be Thinking More Outside the Box When it Comes to Science Funding
In recent weeks and months, a discussion has emerged online about whether the approach the US government takes to funding science is broken and, if so, what to do about it. Maxwell Tabarrok has a good post on this topic, and the issue has also received some attention from notable Silicon Valley types like Patrick Collison and academics like Tyler Cowen and Alex Tabarrok (Maxwell’s father), among others. Many of the reform proposals center around ways funding from government bodies like the National Institutes of Health (NIH) and the National Science Foundation (NSF) should be reorganized, such that the highest-value research is ensured funding. Here, I want to offer a different vision for science funding, one that puts aside the issue of how these existing institutions should be managed and run and instead focuses on an alternative funding paradigm: one built around a sovereign wealth fund.
The younger Tabarrok’s essay does a good job laying out a basic argument for why science funding is currently inefficient. A lot of time is wasted writing and reviewing grant proposals. Moreover, the kinds of projects that get funded tend to be “safe,” in the sense that they are conducted by highly respected researchers from top universities and are on mainstream topics. That may sound good at first, and from the perspective of bureaucrats looking to avoid blame if research doesn’t produce any results, it’s a safe strategy. The problem is the highest return activities can also be the riskiest ones. Given the declining productivity of research in general, there’s a good chance we’ve picked a lot of the scientific low hanging fruit and now we need to go out on a limb to discover the next breakthrough technologies.
Tabarrok offers an interesting proposal for reform: Take the existing level of funding and give it away to the 250,000 or so researchers that receive about 90 percent of funding currently. The logic here is this group would get most of the money anyway, and this would at least save the government the costs of reviewing grant proposals. It would also save the researchers the time of writing them, and these experts would then also be free to use their best judgment about which projects to work on, and this could potentially lead to better research outcomes.
Tabarrok is clear that this proposal is far from perfect, and it’s easy to see why failing to vet projects could mean more silly experiments get funded. But in a way that’s the point. If we can’t distinguish the crazy ideas from the brilliant ones ex ante, funding more crazy-sounding ideas could be just what we want.
My worry about solutions along these lines is that they box us into a corner whereby the current set of institutions and current levels of funding are taken as given. Solutions then end up being about finding the best way to rearrange the deck chairs on a sinking Titanic. My preference would be to focus on two things instead: a) how to make the total level of funding available for scientific research much bigger, and b) how to direct scientific research towards achieving productive outcomes like boosting GDP.
On both of these fronts, a sovereign wealth fund could help. A sovereign wealth fund is a government-run investment body that manages a portion of a nation’s wealth. In the context of funding science, it could be set up in two ways. First, it could fund scientific research directly, which is my preference (although I don’t think it should exclusively fund research). Second, it could fund other kinds of investments, like in real estate or equities, and then take a stream of its annual returns and put them towards funding scientific research, say by giving the money to NIH and NSF. A model here would be the permanent funds in Texas, New Mexico and Alaska, which use natural resource wealth to pay for education, or, in the case of Alaska, to pay for the state’s universal basic income.
The logic here is similar to the “earning to give” strategy the Effective Altruism movement has promoted at times, namely that the best way to “do good” is sometimes to make as much money as possible. If you are a young person exploring what career path to take, and your goal in life is to help others, then, counterintuitively, it might make sense to go work on Wall Street for a decade or two, get rich, and then give all your money away. You could do more good that way than, say, by going to work immediately for a non-profit in India. This career path is precisely the one effective-altruist Sam Bankman-Fried followed to great success. He founded a cryptocurrency exchange, became a billionaire, and is now heavily involved in philanthropy. Unlike a lot of billionaires who discover charity only after they get rich, philanthropy appears to have been a major part of his strategy all along.
Now apply this same logic to science funding. Rather than figure out how to dole out an existing pile of money more effectively, why not find ways to make that pile much, much bigger? In other words, let’s leverage the power of compounding returns to the advantage of science.
In addition—and this is more controversial—I believe the research we fund itself has to produce more tangible results if it is going to put an end to the slow growth that has plagued developed countries in recent decades. Think for a moment about what GDP is: It’s the value of the final goods and services that get produced and traded in markets in a country in a given year. If we want to make GDP go up, it therefore follows, then our focus needs to be on production that is concrete and makes money. Research that makes us feel warm and fuzzy inside just doesn’t cut it.
In the context of science, one way to accomplish this end would be to focus on research that is more explicitly aimed at earning commercial returns. This is precisely what a sovereign wealth fund does—it earns commercial returns—which is why one could have a significant advantage over existing bureaucracies when it comes to selecting worthwhile research projects (at least if we agree our goal is promoting economic growth).
This need not come at the expense of the public interest either. Research on vaccines, for example, has been irreplaceable in the fight against the coronavirus, but it has also made pharmaceutical companies filthy rich. A sovereign wealth fund could allow the taxpayer to participate in this bonanza, whereas currently she is left on the sidelines while others get rich at her expense.
A sovereign wealth fund could take ownership stakes in patents resulting from scientific research, giving it and the researchers involved both skin in the game. Another model is the way that universities earn licensing income when research developed in a campus lab ends up used at successful spin-off companies.
A concern that might be raised about this strategy is that lots of research creates value for society but doesn’t make any money. No doubt this is true, but setting a minimum required rate of return for science funding is at least one way to separate the wheat from the chaff. Moreover, taxpayer funded research should arguably be held to a higher standard, not a lower one, than private funding. If people want to spend their own money on research that serves no practical purpose, then by all means that is their right to do so. But this is not, in general, how we should be thinking about spending the taxpayer’s dime.
I do believe some leeway should be granted when it comes to how high on average required rates of return should be. If a portfolio of scientific research generates some positive annual return, say 6 percent a year, as opposed to the very maximum rate it could possibly achieve, I see this as okay. Some of the portfolio could constitute investment viewed as charity or sufficiently long run in nature that it takes awhile for it to pay off. Moreover, not all social returns can be captured privately, so it should be recognized that some investments might generate social profits, even if they generate a private loss on the government’s balance sheet.
However, I view it as extremely problematic if an entire portfolio of scientific research produces financial losses on a recurring basis. Just as the economy on net cannot afford to have investment as a whole produce negative returns in perpetuity (nor, for that matter, can any individual private business), publicly funded scientific research should be held to a similar standard.
In addition to creating wealth for funding scientific research, there are many other advantages to sovereign wealth funds, which include promoting economic efficiency and intergenerational equity. Those interested in these topics can read my previous writings on these topics elsewhere. In addition, a US sovereign wealth fund could potentially help address global savings imbalances and reduce wasteful domestic government spending and debt. (My colleague David Beckworth and I agreed to write a paper on this topic a while back, though I keep getting side-tracked with other projects so it could be a while before it comes out.)
Tyler Cowen recently opined on Twitter that little discussion has been stimulated in the academy about the merits of science funding reform. That’s despite the best efforts from him and others to start a dialogue. It’s good therefore to see that some on the internet are starting to talk more openly about this issue. However, I worry the current conversation is too narrow. We should be thinking more along the lines of shifting the paradigm for science funding and less about how to rearrange an existing set of institutions that may not be fit for the 21st century.
The problem with a sovereign wealth / impact investor fund is that you still need to invest in projects that are profitable, but may not actually maximise public good, because public goods lack a business model due to inability to leverage IP or private property rights to make a ROI because they are non-rivalrous and non-excludable.
It would be incredibly powerful to combine this concept with a Social Impact Bond (SIB) or Advance Market Commitment (AMC) (also referred to as Pay-For-Success (PFS) contracts or Retroactive Public Goods Funding) to incentivise impact investors to fund public goods directly. If the public good saves the government money or exceeds the value of what they would be willing to pay, then this would be scalable as a business model (e.g. $10k per QALY generated by repurposing an off-patent drug, which is cheaper than $30k per QALY paid for patented drugs / homeless person housed for a year / x tonnes of carbon removed).
We are interested in pursuing this SIB / AMC to incentivise the development of public good medicines which are unmonopolisable using traditional IP (e.g. generic ketamine to treat depression or fluvoxamine + inhaled budesonide to treat Covid, nutraceuticals, plant medicines, psychedelics, diets, lifestyle interventions etc).
If interested, please reach out at crowdfundedcures.org
James, I think your thinking here is more or less along the lines of our thinking at The Lianeon Project. Since patents are the creation of the state, we envision partially publically owned patents, with recurring taxes placed on that ownership share, probably around 5 percent annually. The funds raised by this tax are used to buy out IP that would better society immediately, and some diverted to a DARPA-like organization that complements, but does not compete with, private sector R&D.
Read it here: https://www.lianeon.org/p/turbocharging-technology
Would love you to join us at The Lianeon Project