Dec 16 2010

Should We Democratize Science Funding?

A recent opinion piece in the New Scientist by Dan Hind continues the debate about democratizing science – changing or adding to the institutions of science to introduce more democratic public participation. This is an interesting debate, with several facets, and I have some mixed feelings about it.

Hind’s article is mainly about science funding, which I will get to shortly. But if you search for articles on “democratizing science” you will see that the issue extends beyond funding. For example, it also includes the notion of making journal articles available for free to the public online. I am completely in favor of this form of democratization – open access, all the way. I understand the need for journals to be viable commercial entities. I would like for journals to find business models that are viable but still allow for open public access to content.

I am spoiled because I have institutional access through my job to most journals, but when I am away from work or want access to a journal that Yale does not subscribe to, it is extremely frustrating not to access to the full articles I need for my research. Sometimes this affects my blogging (and therefore must also affect other bloggers), and therefore it detracts from the public discourse on the topic. Usually the articles are available for pay – at what I think are ridiculous prices, often $50 or more for one time access to a single article. That’s just not in my blogging budget.

This is why I applaud and support journals like the Public Library of Science (PLOS) – which is a series of online journals with open access. But other journals need to explore more accessible options as well. Perhaps, for example, journals or groups of journals (grouped by publisher or topic area) could offer low cost subscriptions to the general public. There are lots of variables to experiment with – access to some articles, delayed access, access to articles but not certain special features, etc. Professionals and institutions may still pay more for premium access of some sort. Or journals may just need to find some other revenue streams to keep subscription costs accessible – keeping in mind the public (unlike professionals) don’t want to read every edition cover to cover – they may just want access to a single article once ever.

So, with regard to access to scientific articles, I am in favor of open access, democratization, and experimentation with business models that allow for reasonable public access.

Funding of scientific research, however, is a different issue, and that is what Hind is specifically writing about. He suggests that part of public funding for scientific research can be allocated by a democratic public vote, rather than panels of experts. Here is the crux of his argument:

While this research has undeniably delivered important public benefits, it has also tended to drive “the whole thrust of the economy” in directions that favour powerful elites. The profits that derive from taxpayer investment in science have overwhelmingly been captured by a handful of investors and senior managers. Furthermore, a system of opaque subsidies gives favoured sectors of the economy access to vast sums of public money, which they use to develop products that the public would not necessarily wish to fund. The pharmaceutical sector, for example, has spent billions on copycat drugs and treatments for depression and anxiety that have few clear benefits.

There are two main points to address – are there really such downsides to the current system of funding, and would a democratic system be any better? Keep in mind, he is advocating allocating just a small part of public funding to a democratic system, not all of it. In any case – I think his characterization of the current system is not entirely fair.

First, I would point out that there already are multiple sources of funding for scientific research. Public funding is by far the largest source, but there is also corporate funding. Corporations fund research with their profits (although sometimes also with government subsidies which the government uses to promote research in desired areas). Commercial research has the specific goal of leading to marketable products and services.

It is difficult to argue that such commercial research is not directed toward products the public wants – that seems to be the whole idea of commercial research, to develop products the public will want to buy. I understanding that marketing is often about getting the public to buy things they don’t really want or need, but there are limits. In general, if you want a killer product, make something the public will find useful.

I also find it interesting that his main example is copy-cat drugs by pharmaceutical companies. I find this to be a lazy example, and perhaps the most often cited one I hear. There is a role for copy-cat drugs. They provide more options for clinicians. Patients may have allergies, or not tolerate the side effects from one drug, but respond well to a “copy cat”. So while they are not as useful as an entirely new class of drugs, they do provide added benefit to treatment.

Further, pharmaceutical companies invest billions in risky research. Many do not survive long enough to see a profitable drug to market. So I have no problem with drug companies putting a tried and true product on the market, even if the benefit is incremental, to fund their research into more risky new drug development. There is downstream benefit to drug companies funding their operations with copy-cat profitable drugs. This is the marketplace in action – which is ironically what Hind is primarily arguing for.

It should also be noted that any researcher who thinks they are on to something can raise venture capital, or create a startup company, and fund their research. All they have to sell is their idea.

In addition to corporate research, there are also private and non-profit advocacy groups that fund research, usually with a specific agenda in mind, such as curing a specific disease. Charities raise billions from the public, who vote, in essence, with their donations, and then fund research the public wants.

With so many options already available, do we need to allocate public funding for scientific research to a democratic process? I don’t think so. The government has the responsibility to spend public money responsibly. We live in a republic, not a true democracy, and this is an area where I think the former model works. The public pays taxes and trusts that their money will be put to good use for the public good.

I have argued before that research resources should be allocated by a non-partisan process of appropriate experts evaluating the potential of research on purely technical grounds. Politicians should resist the temptation to put their fat thumbs on the scale, which inevitably results in ideology trumping good science. Hind argues that this is unavoidable, but I disagree. We just need to keep the pressure on, to let the scientists do the science, and criticize politicians who think they know better than the experts.

I also don’t have much faith in what would emerge from an open process of voting on what should get funded. This may sound elitist, but it is just a concern for quality control and appropriate standards. If we have learned anything from the process of online voting it’s that such open feedback is easily distorted. Special interest groups could have a disproportionate effect on public feedback. The more fanatical or fringe the group, the more likely they are to be highly motivate to advance their agenda.

Think about it – won’t this just result in creationists voting for creation research, anti-vaxers voting for dubious and unethical research on vaccines, every charlatan campaigning for funds to study their quack remedy, and basically everyone with an ideology or fringe belief vying for public funding of their pet project? I have seen enough internet polls to know this is what basically happens. This would not be an efficient or responsible use of limited public research resources.

The underlying assumption of Hind’s position is that the public has a reasonable sense of where research dollars should be spent. But let’s put this assumption into context. By any reasonable measure the public is largely scientifically illiterate. Science education is widely criticized as being inadequate. The press does a good job of sensationalizing science and giving the public a very distorted view of science and research – often emphasizing minority, fringe, and even outright pseudoscientific ideas.

Democratizing science, in the end, may just be another way of politicizing it. If any significant funds were allocated by a democratic vote, then politics will enter the fray. The history of politics kibitzing science is not a good one.

I also challenge the very notion that popularity is a good measure of what should be researched. That is often the justification used for institutions like the National Center for Complementary and Alternative Medicine (NCCAM). The NCCAM, in a way, is an experiment in allowing popularity to set a research agenda. So far they have spent 2.4 billion dollars of taxpayer research money, and have no positive findings to show for it. In fact there is evidence that popularity predicts low reliability in research – it is actually a very bad indicator of which research is promising.


I know that the notion of democratizing the allocation of resources resonates with our cherished sensibilities. This is precisely why movements based upon the notion of “freedom” or condemning “elitism” work – it’s an easy sell to a society that cherishes individual freedom and self-determination. I share those sensibilities.

But some functions of society are too complex to be left to the vagaries of inexpert opinion. For example (this is perhaps a more intuitive example), would you want the fate of those accused of a crime to be determined by a popular vote? Do you think this would lead to greater justice? There are certainly problems with our judicial system, but the system does basically work. At least there are rules of evidence, procedures to protect rights, and ultimately judges, who are experts in the application of the law, to oversee the process. Democratizing the courtroom is akin to mob justice.

What Hind is arguing for, in essence, is mob science. Decisions about what research avenues are promising are at least as complex as applying the law. Both processes are messy and imperfect, and even riddled with biases, but they basically work. The flaws are no reason to favor a wild-west type approach (that’s throwing the baby out with the bathwater).

But I also think there may be a kernel of value in Hind’s suggestion, in that I would like to see the public more involved in the process of science. Some have argued, and I largely agree, that non-experts should content themselves to be cheering on the sidelines of science. It takes years of study, and painfully acquired expertise, to make meaningful contributions to a complex and advanced scientific field. That is not elitist – that is reality. It is no more elitist than saying that an out-of-shape amateur has no place playing in the major leagues.

This, however, is a dilemma of modern science. It is really complex, so the public cannot directly participate in it, but we do not want the public to feel isolated or mystified by modern science. So we need to find compromises in the middle. The is the goal of all attempts to popularize science – to make it accessible, fun, and interesting to the general public.

Project like SETI at home are ways for the public to play a little part in a larger research project. Interested individuals can also identify galaxy types, or practice folding proteins online. These are all great project that allow for public participation in science. But at the center of all these projects are scientists who know what they are doing.

Perhaps Hind’s notion can be altered to fit this model. The public can be like the jury in a court case – they get to vote on what they would like to see funded, but they don’t get the final say. Their vote becomes one factor of many that is considered in allocating funding (along with plausibility, practicality, utility, and ethical considerations). Also, make the process interactive, so interested members of the public don’t just do a drive-by vote where they dump their biases or fringe agenda then move one. Perhaps there can be a moderated discussion with input and feedback from scientists.

The public can post their ideas or vote on their recommendations, but then scientist can provide input such as – that idea has already been tried (here’s a reference), or that is not practical for this reason, or perhaps such a study would be unethical. Who knows – good ideas might emerge from such a process. Scientists may learn more about what the public thinks about their specialty, and this could be a good opportunity for scientists to explain why they research what they do (why do scientists, for example, find fruit flies so fascinating).

In short – public participation in science is good, but public determination of science is hugely problematic and counterproductive.

In the end we have to remember – science is not a democracy. It’s a meritocracy.

16 responses so far