A Fundamental Critique of Scientific Funding Now and Future

Reviewing DeSci’s Mechanisms

Introduction

This article explores the funding mechanisms in traditional science, and Decentralized Science (DeSci) and presents an analysis of a gap to be filled.

1. My Suspicion About Revolutions
2. State of the (Scientific) System
3. The DeSci Funding Stack
4. Who is Doing the Work

A quick note on the scope: I’ll focus here on scientific funding.

I do this for several reasons. First, I am a social science undergraduate. I see almost everything as a network through which we transfer relationships. Money is a neat and traceable relationship — and one that, in our current paradigm, is a good proxy for power. In science, if I have money, I get to share my ideas. If I don’t… I don’t get to share those ideas. Second, how we manage and distribute money is a close analogy to our worldview. I want to know — is our new transition really a transition? And if it’s not, what is missing?

Generally, I believe in the toolset of Decentralized Science. I think we could use it to distribute and improve scientific benefits and research. I also believe that we can use it to reinforce the same dynamic we’re trying to fight against. I’d like to approach the outcome of the first.

This article is a precursory analysis and critique. Please comment and deride :)

Science in Transition

Scientific funding is facing a public crisis. The funding situation has evolved from royal and religious patronage, to funding through institutes, and now science is largely funded through a ‘market-of-ideas’ model where governments or corporations evaluate research proposals before funding it.

This model creates an arbitrary set of rules, by arbitrary groups, with arbitrary (and costly) requirements that defines what type of science gets done.

Recently several movements have come to the fore to resolve a huge number of these limitations, and one of the sticking points has been funding. Open science wants to throw out the journal-profit model in favor of open access, but at the moment pushes the cost onto individual researchers for publication. And doesn’t change the fundamental problem of how much it costs to get funding, and the bias that emerges from who funds the research.

Another movement — DeSci or decentralized science is attempting to resolve this dilemma. The movement aims to shift the labor-heavy, one-off, and biased or arbitrary scientific funding to a low-effort, sustainable, and more objective method of determining who should receive funding.

Core to measuring the success of this movement is achieving Feyerabend’s principle of ‘against method’. Right now private corporations, journals, and the government are the only ones to determine what can be researched because they hold all the funding in the relationship. The goal would be to create a system where we can widen scientific discovery by making sure that anything can go.

The Efficacy of Revolutions

I like the framing of DeSci. But I’m skeptical of the new science ‘revolution’ because revolutions are questionably effective.

Many root themselves in the claim that they’ll bring power to those who don’t have it. While some seem to achieve their goal, others fail epically, and others eliminate the original cause of inequality only to replace it with another.

This pattern is especially pervasive in science and technology. New innovations arise boasting of the great improvements we will achieve in quality of life, only to reinforce a pervasive inequality that seems to improve our lives only an average but not absolutely (Lou Keep has a nice essay on this — he comes to different conclusions than I would but covers quite clearly why increasing GDP simplifies a socially unstable world into an unhelpful metric).

In “Winners Take All: The Elite Charade of Changing the World,” Anand Giridharadas hypothesizes that this pattern will only be reinforced.

Change right now is driven not by those most hurt by the current system, but by those who have benefitted the most. The perspective of the global rich is that they are destined to drive change through their wealth.

The fundamental limitation of this model of change is that the most well-off have won. The system has served them. As a result, the way that these groups model creating change is through peddling and reinforcing the system that set them on top in the first place. Not in any evil-self-serving-die-scum way, but through the simple mechanism of demographics.

Boaventura de Sousa Santos would call this an Epistemology of the North. The lived experience and constant validation of our current economic system mean that what is seen as valid by this elite will more likely than not fall within the system that exists. If we see any system with winners and lowers as a big game, then the winners of the game have been rewarded again and again by believing in the game.

Any new system they create will probably look like what they think works.

On the other hand, de Sousa Santos aimed to elevate the ‘Epistemologies of the South’ as a way of reclaiming the validity of the lived experiences of groups who have not been served by our current resource distribution system. In our game, these are the groups who can see the ways that losing once makes you more likely to lose, or other limitations that the winners are less likely to see. Or alternatively — they’re the groups that never participated in the game in the first place.

In a revolution driven by those with extra resources, the people who haven’t participated or haven’t been served in our system are — by definition — excluded from the subset of change makers.

In this frame, revolutions built by the winners are questionably effective. And this joint frame gives us an idea of how to evaluate if a revolution might fundamentally improve the situation.

1. Who are the revolutionaries? If they’re made up of a single group of winners, then we’re more likely to have reform than revolution. Reform isn’t bad necessarily — but at the very least, it has cut out some percentage of potential solutions by only hearing solutions from one side of the story.
2. Do the mechanisms created by these revolutionaries represent a departure or replication of the previous system? While replication isn’t inherently bad, it’s more likely to end up with very similar problems to the original system we’re trying to change.

What is our Current Funding System?

The current knowledge pipeline in scientific research looks like the figure above. Universities determine who the potential knowledge creators are, and largely how that knowledge gets created. Governments, universities, and corporations get to determine what gets made by choosing who gets money for their work. Journals then evaluate that knowledge and determine what of the knowledge that is created gets published. And then Journals also get to determine who gets to access and use that knowledge by creating article fees and other ‘barriers’ to the work. This means that largely — academics get research through the access provided by their institutions. And others can only get access to papers that are sourced openly or the ones that skip the journal filter altogether.

We’ve only got corporate or government interests to fund science (not the science itself). For corporations — the motivation is largely profit. Will this research, if true, justify, support, or advance my capital position in the market? This limits the research that is effectively funded by this mechanism to applied research.

For governments, 40% goes to defense, 27% goes to health and human services, 12% goes to Energy, and 8% goes to NASA. The rest is allocated for everything else.

The method to get this funding is typically a huge grant process in which researchers spend up to 40% of their time on securing money. Worse, the agreement between grant reviewers about if a grant is a high quality is shockingly low. It doesn’t seem like the decision about what should be funded could be considered reasonable. In addition, only 20–25% of grant applications are awarded. This means that there is lots of time spent thinking up and fleshing out ideas that die in the file cabinet untested.

The alternative is getting hired in corporate research where your funding is secured, but your outcomes are controlled.

The core critiques against this system have been levied by DeSci and Open Science as such.

1. This is not a sustainable way to fund science. Scientists can never reduce the amount of time they spend on convincing grant foundations to give them money. Thus researchers will always be dependent on the structure.
2. This work is an opportunity cost — researchers spend less time doing research if they’re spending more time speculating about if they can do research.
3. This funding process puts the control in the hands of a few grant panels and corporations who relatively arbitrarily define what research should or shouldn’t be funded.

This current inefficiency, dependency, and bias limits scientific discovery both in terms of the amount that can be created, and in the types of science that is done. In our current system, we create a systematic bias against basic, long-term, or social discoveries or knowledges. There is no place at all for knowledge that hasn’t been presented as ‘science’.

In general, systematic biases lead to more blind spots.

Who are the winners at the global level?

• Corporations
• Government

Who are the winners within the academic system?

• Well-funded scientific fields: Defense, Applied Research, Health Sciences.

What is the mechanism of the system?

• Anticipatory idea market. Funding is more likely to be applied to those who have been successful in the past.
• Capital market. Funding supports those researching the things that corporations want.

The Problem

From above, we see a leverage problem. Neither researchers who are closest to the research nor the public who will benefit from it has any autonomy to determine the type of research that gets done. The only science that gets done is research deemed fit by the government, corporations, or journals.

Why did these systems emerge? What do they claim to do? The government grant system emerged in science as a first step to move science away from pure patronage. This opened science up to all those who wanted to participate in institutes or universities (provided you met the requirements of the institutes) instead of those who were funded only by the wealthy.

The patent system for science similarly emerged as an attempt to transfer scientific insights directly into use for the general public. This aimed to eliminate the barrier between scientific discovery and its impact on the public.

While we can argue that the evaluation techniques for grant proposals and the use of patents are weak at best, both the current systems raise a point.

Money is scarce — we need some way to know how to allocate it.

The DeSci Proposal

DeSci looks at the current problem and sees that right now we have very few very controlling middle men between researchers and their research outputs. As part of the wave to decentralize many different industries through the Web3 movement, DeSci aims to free researchers and research to self-determine by shifting the power from these institutions into the scientists and the public. At the same time, it hopes to continue to do the service of knowledge validation and capital allocation that governments and corporations currently do.

This new system aims to coordinate the creation, evaluation, and use of knowledge like so.

In this system, DeSci says everyone gets access to scientific research and anyone who wants to can enter into science. Similarly, the barrier between being a potential knowledge creator and being funded to create knowledge is other scientists and the public.

But this leaves the core problem. If everything is open — where is the potential for profit? And where does the funding come from?

DeSci attempts to resolve the dilemma in several ways.

First through Decentralized Autonomous Organizations (DAOs). A DAO is an online community of people who have defined voting and other coordination within the community on a blockchain. This means they have carefully defined how each individual can support and define how the community uses joint resources. In this model, the DAO is funded through the community who joins.

In DeSci, they go beyond a simple one token one vote system in which a single actor with lots of capital can buy many tokens and uniformly determine the community outcomes. Instead many of these communities use a quadratic funding mechanism. Quadratic funding claims to be the optimal way to allocate funding for public goods. It does this by considering both how much a member gives to a certain good and how many members give to that good.

This works in two ways to avoid the tyranny of the wealthy. First — if one person votes many times for one good and two people vote fewer times for another good. Than the second good will get more of the collective funding. This means that the more a wealthy actor wants to place on a single good, the lower the impact of their votes. Thus goods with more popular support will be funded, and wealthy actors will be incentivized to spread their votes across many different goods ensuring a more balanced funding ecosystem.

Another mechanism developed for both grant makers and DAOs alike is retroactive funding. Retroactive funding gives funders the ability to fund projects based on the impact the projects have already had. Thus as a project increases it’s citation count over time, is used by other members of the community, or is otherwise evaluated as credible, it can be funded in an ongoing manner. This way the researchers who do good research can keep doing this good research without forcing themselves to write extensive proposals in order to convince funders about the relevance of their new research.

Finally, for applied research, IP-NFTs hope to provide traceable benefits to the research that was transformed into capital outcomes. By placing IP on an immutable blockchain we eliminate some of the bureaucracy around IP markets, provide a traceable link between research and corporations, and can automate royalties or other benefits to the researcher.

What Are These Mechanisms?

To parallel the structure above I will define a) who has power in this system and b) the mechanisms that determine who will gain and maintain power.

I’ll start with b.

DAOs are markets.

IP-NFTs are markets.

Markets are great. But they have core limitations. As indicated by the improvements in DAOs made using quadratic funding and retroactive funding what we can clearly see that the movement is attempting to resolve some of the core issues with markets. In a relatively distributed system, Quadradic funding limits the snowball of capital aggregation that is known to create monopolies and thus high barriers to entry in these markets. Retroactive funding makes the information we’re deciding on much clearer.

Speculation is almost always 50/50, but hindsight is (in theory) 20/20.

However, the fundamental nature of the market is unchanged. We can distribute goods globally in a way that any individual only needs to act in their own best interest. Best interest is essential and useful, in the case where ‘interest’ is clear to the individual, and where ‘interest’ is now. In science, this is far from true.

We see this in our current system in how basic science, social science, and knowledge evaluated on different terms are essentially eliminated from the corpus of knowledge through lack of funding and support. That doesn’t make this research useless or unworthwhile.

Basic research may be applied and later, or it may never be applied. It is both a gamble and its a gamble that we won’t get feedback on until far in the future. Markets don’t like gambles and they don’t like future payoffs because market actors are risk averse and short-lived. And while the claim of markets is that there is a niche for everyone, a niche of excited chemists supported only by other excited chemists is unlikely to gain much wide spread support. Even in a quadratic system they don’t have either the numbers or the capital to unilaterally support themselves.

Similarly, knowledge that isn’t marketable has limited impact for the self-interested. Research about how to live well, preventative measures for diseases instead of paid for cures, or information about other ways of living and relating are free bits of information that is worthwhile, but in a scheme of putting in tokens as an investment based on how well you think a proposal will do in a market it’s unlikely to battle out the next big drug discovery.

Retroactive funding further fails to solve this problem. At very least it fails to do so in the short term. A researcher whose work aggregates citations slowly and affects only a small group of chemists will likely not be sufficiently funded to continue to work. The basic researchers still need to fight for funding from somewhere. We create another feedback loop — applied scientists can keep creating after one success, basic researchers are still caught in an inefficient grant system that saps their time.

Finally, IP-NFTs exacerbate this problem. Once again, a single success sets an applied researcher up to continue making more successes. Even within a market of other applied researchers, this system reinforces the dynamic of those who succeed will keep succeeding relatively independent of their subsequent success.

We’ve criticized the biasing effect of a single actor. But the mechanisms we have have recreated a bias towards applied and immediate science.

—

None of these mechanisms are bad. None of the problems I brought up are unsolvable. There is even a chance that they could all be solved by a very very very carefully crafted blockchain protocol that considers all the incentives so we make a robust scientific ecosystem.

But if we dig into the first half of why revolutions don’t end up being so revolutionary, then I’m not that optimistic that we’ll start crafting mechanisms that are able to adequately balance these different types of knowledge.

DeSci is largely made up of people in Bio-med, Bio-tech, Metascience, and Web3. Besides the metascience people, all of these people are clearly served by the market model we have today. Yes — they can see the problems and the limitations of science, but the market mechanism still largely works for the clear and quick outcomes of these fields.

The mechanisms we’ve created thus far don’t support basic scientists, social scientists, or anyone else because these people largely don’t seem to exist in the DeSci ecosystem.

—

Part of DeSci aims to adhere to Feyerabends argument in ‘Against Methodology’. Anything should go — science isn’t always transparent, we don’t always know what should be funded, what will be big, what will become essential, or what will ultimately help us. Exploring as widely as possible expands our options for hitting on what’s truly ‘important’.

Right now — we can’t claim that anything goes in the DeSci ecosystem. Largely because, there is no ‘anything’ when we all agree on the one thing. And our mechanisms reflect this — right now they support the one thing.

So what’s next? I don’t know — but I propose a gap. We haven’t yet started building mechanisms that fundamentally solve the problem of short term, market-based, applied science, bias. I don’t think we will until we have basic researchers working on funding. Indigenous scientists working on funding. People from the Global South. Psychologists and political scientists. And everyone else for whom bias means the difference between contributing to science and not contributing to science, and not simply the degree of funding you will get.

It’s time for funding in DeSci and Science in general to be a conversation. Not an agreed upon claim about ‘optimal’ allocation.

It’s going to take some work.