Is science bad for politics?

Daniel Sarewitz argues that yes, it is, and we're better off first hashing out value disagreements first. But shifting from arguing about facts to arguing about values is equally difficult.

Note: Daniel Sarewitz’s “How science makes environmental controversies worse” is the inspiration behind this assignment I use in my class.

Often, it is in graduate school that social scientists read papers that stay with them forever but I read Daniel Sarewitz’s “How science makes environmental controversies worse” not in grad school but in my first—and very difficult—year of teaching. I was thinking about readings to put into my “Science, Society, and Ethics” course and it was listed in one of the sample syllabi I found on the web. The title seemed interesting, and when I read it, the paper blew my mind since it seemed to synthesize what I had learned about science, society, and politics in crystal-clear language. (I can’t begin to tell you how well-written this paper is; no line seems superfluous, it moves towards its climax superbly; and there is hardly any jargon.)

That’s enough praise. What’s the paper about? Sarewitz argues that we cannot expect science to solve political controversies. This is not because one side is anti-science and refuses to accept what the science is saying, a common belief among some left-leaning commentators. It is because any side in a big technical-political conflict (such as whether to build a nuclear plant, regulate GMOs or fracking, or to shut down schools during COVID) can enlist science to its cause.¹ Science is multi-faceted; it has at its disposal different disciplines, different methods, and different funding structures. Sarewitz argues that when we shunt a conflict about what we need to do towards science, when we say, “let’s do the science and it will tell us what to do,” we don’t resolve the conflict; instead, we funnel it into science itself. Which is to say: more political stakes, when channeled into science, lead to a lot of science, but this ends up increasing uncertainty rather than reducing it.

Sarewitz’s solution? He says that the best way to solve technical-political controversies is through good old-fashioned politics. Politics is the essential channel to solve and resolve disputes over values and policies; science should come in after political negotiations. As he puts it, “it is only after values are clarified and some goals agreed upon that appropriate decisions about science priorities can emerge.”

I love this paper (and it inspired the assignment I created for my class). But I think one of the problems with this paper—or maybe how it is interpreted by its readers—is that it makes it sound as if problems are easier to solve politically. But if science is difficult, politics is even more so. What Sarewitz is suggesting is not that we simply fight harder in our political fights but that “political innovation²” is just as important as scientific or technological innovation. We need to design and redesign our institutions in which we conduct our politics in order to both accomplish goals (although these goals are obviously contested) but also to protect science itself.

Why does science not resolve political controversies? Why is this bad for science and society?

The political stakes of a debate are directly correlated to the number of scientific and institutional players which, in turn, is directly correlated to scientific uncertainty. When the stakes are lower, there are fewer scientific and institutional players, and the uncertainty is ALSO lower. When stakes rise, we get more players, therefore more science, and thus—and this is Sarewitz’s main point—more, rather than less uncertainty.

Sarewitz’s key point (summarized in the figure above) is that more science, funneled by political stakes, leads to more uncertainty, not less. Why is that? Because now everyone with a stake in the conflict starts to do science; findings that may have been agreed upon before suddenly become contentious as many more bring their skepticism to bear and start to deconstruct these findings. Their methodologies are questioned; their unit of abstraction is deemed to be wrong; as the philosopher of science Bruno Latour would say, the black box of the finding is pried open and all its contingencies spill out for all to view. Even the motives of the scientists become objects of scrutiny and questions like: Who’s funding this scientist? Who funded this scientist in the past? Are they shills for so-and-so industry? Are they just unabashed leftists? Have they done something in their past that casts a shadow on their honesty and integrity? all suddenly start to come to the fore.

And here’s the thing: on any, big complex topic, the science has many different facets; it is full of cracks, and those cracks, in the wake of relentless scrutiny, threaten to become gaping fissures and fault lines. As a result, we get a lot more science, but also a lot more uncertainty. When political controversies are deflected on to science, Sarewitz says, the controversy remains a controversy and if anything, by scientizing the controversy, we have now politicized science.³

The pro and anti GMO sides in science vary with respect to their disciplines, values, methods, and interests. It seems highly unlikely that they will converge on to clear factual claims about the risks and benefits of GMOs in scientific terms.

Sarewitz is at pains to point out that none of the scientists fighting it out are doing it in bad faith or are doing bad science.⁴ In fact, the political stakes help make sense of the arguments.

The paper’s best example is GMOs. Are GMOs “good” for humanity? It turns out that this depends on what you mean by “good.” One pro-GMO group of scientists turn out to be molecular biologists and plant geneticists who argue that there are many benefits to GMO technologies because they can help us design plants with greater crop yields and greater resistance to pests (which in turn can help reduce pesticide usage). On the other hand, we have scientists, often ecologists and population biologists, arguing that the problem with GMO technologies is that we don’t know what they will lead to; e.g., do the new genes find their way into “natural” plants in the wild? How will they affect non-pest insects like bees?

As Sarewitz points out, there are hidden values in these debates. For instance, the molecular biologists and plant geneticists are making an argument for GMOs based on their intended outcomes while the opposition is focusing on unintended consequences. This is a value difference in that it is an attitude one takes towards risks. But their differences don’t end there. They also use different methods: molecular biology is characterized by highly controlled lab-based research while ecological research is field-based; with methodologies that are so divergent, it is hard to reach agreement. And finally, they have different interests: molecular biologists are much more likely to work in an agricultural MNC while the ecologists are much more likely to be affiliated with environmental groups.

Ultimately, then, Sarewitz says that the resolution of the GMO controversy will not happen because one group of scientists will refute the other. It will only get resolved once a political settlement or compromise is reached on many of these underlying questions that are not scientific at all. For instance, how should we think about what we prioritize? Are we okay with more food production in return for less biodiversity? Are we okay with the business models of the agricultural MNCs? If we focused on these questions, Sarewitz says, we could come up with policy options that can lead to compromise (and open up more paths down the road). The science would help us with these policy options but it would not be weighed down with the responsibility that it alone should dictate the policy options.

Sarewitz says that this scientization of political disputes is bad for two reasons:

(1) It is bad politics because it prolongs the conflict without producing any action. Action, for Sarewitz, is the most important when it comes to decision-making and the constant scientific conflict might give an illusion of action when there is none. In a partisan sense, endless standoffs and no action are bad for progressives who are often the ones fighting for change. But an issue that never concludes in either direction is just bad for everyone; better instead if a decision were to happen, albeit one you disagree with, so that you can fight another day.

(2) It is bad for science because it reduces the public’s trust in science in general. The public can see for itself the grubby nature of science, as experts with PhDs working in government, corporations, and NGOs fight amongst themselves in the most take-no-prisoners way imaginable. Even if you favor one side in the scientific dispute, this spectacle does not do the enterprise of science, as a whole, any favors. Just think: if these experts can fight about everything in this way, can we really trust them with the other things they say and do?

Is politics enough to resolve political controversies?

Okay. So what’s the solution? Is there one? Sarewitz says that it is better instead to leave political conflicts to politics rather than shifting them into science.

Why? Because there are established channels of political negotiation and value trade-offs and these are deemed broadly legitimate by most people. On the other hand, there are no rules for value negotiations in the scientific process which is premised on the idea that science is value-free (or at the very least, that its values are non-political or embodied in methodological considerations). By definition, when a political fight is fought in the scientific arena, the scientific actors are primed to hide their values and fight only on the basis of facts—and the facts, of course, are too complicated for a simple resolution. Thus:

Political debate permits the mobilization of a broad range of weaponry, including scientific facts, religious dogma, cultural norms, and personal experience, in defense of one’s values and interests. But scientized debate must suppress the open discussion of value preferences; were it not to do so it would have no claim to distinction from politics.

Broadly, I agree with this: it is better to have a political debate about policy options than it is to have a scientific debate about facts in the hope that it will settle the policy once and for all.⁵

But this does not mean that political negotiation is easy or that it will yield results. Political negotiation is just as likely to get stuck, just as likely to result in stand-offs that result in no action. Moreover, political negotiation is just as likely to lead to policy tradeoffs that progressives find distasteful.

Perhaps that is why the best gambit (one might say most trollish) in Sarewitz’s paper is holding up the Supreme Court decision Bush vs Gore — which handed George W. Bush the presidency of the United States — as an example of a successful political resolution of a controversy. Progressives, of course, and with good reason, loathe Bush vs. Gore.

The details of the 2000 election fight are too convoluted to get into here but as a summary, it’s sufficient to say that the 2000 US Presidential election really came down to who won Florida and Florida was incredibly close. For a scale of how close, it was ultimately decided by 537 votes (0.03% of the votes cast; for comparison, the narrowest margin of victory in the quite close 2020 election was Georgia with 0.23% or 11,779 votes, this was 8 times more than Florida).

Because it was so close, the question became how to think about all those ballots that are discarded because it is hard to discern the intent of voter. Voters often cast ballots that were hard to decipher (see the figure above on the definitions of the “tri-chad,” “swinging door chad,” “hanging door chad,” and the “dimpled chad”) and that question, usually a technical question, solved through the embodied knowledge of poll workers, became an incredibly contested political question because now those hanging chads became symbols of what voters wanted.

Sarewitz argues that if we had tried to resolve the Florida vote question through science, it would have lead to an endless cycle of contested and re-contested findings, partly because so many interpretations can be put onto the meaning of the ballot by different scientists who worked with different methods, had different values (should elections focus on voter behavior or their intent?), and also different interests (they favored Bush or Gore). It is true, he says, that “uncertainties that were revealed in the Florida election no doubt exist in all elections,” but they became important because the election was “so close” and “the stakes were enormously high [because] the presidency would be determined by its outcome.”

If a scientific resolution of the Florida election dispute would have been an endless fight without resolution, the actual political resolution was ultimately swift and decisive. As the two sides filed lawsuit after lawsuit over a month, the argument finally ended up in the Supreme Court and the justices ruled against (by a wafer-thin 5-4 majority) the Florida Supreme Court’s decision to do a full manual recount of certain counties. That settled the issue as the Gore campaign conceded. While “a purely technical approach” would have become mired in endless disputes, the political resolution worked because it was focused NOT on “arriv[ing] at [the] “truth” but to transparently negotiate among competing players.”

I recently listened to a 3-epsiode series on the Reflector podcast that tried to provide some historical context on claims that denied the legitimacy of elections (culminating, of course, in President Trump’s claim that he won the 2020 election, his refusal to concede in spite of having no evidence of any actual voter fraud, and the horrors of January 6th).

I was struck, however, in listening to the podcast by how close the 2000 election was and how it was settled not just by the Supreme Court’s somewhat arbitrary decision but also by Al Gore’s exceptionally graceful concession speech, a concession that was rooted in the norms of American politics. What if Gore hadn’t conceded? While the possibility is difficult to imagine, Trump’s behavior after the 2020 election provides a kind of counterfactual on what might have happened.

My point here is that it is wrong to assume that if we privilege the politics over the science, the issue can be resolved. After all, the reason that politics was scientized in the first place was that people thought that science, being objective and limited to certain kinds of arguments, would lead to issues being resolved unlike politics, where it seemed like entrenched interests would just fight out a trench war that ended in stasis.

The real conclusion from Sarewitz’s paper is not that politics can help us resolve our problems but that we need to think creatively about the art of politics itself. To solve our technical-political impasses, we need inputs from experts (i.e., scientists), stakeholders, and other political actors. We need to both create a space for scientific findings but make sure that the debate is not scientized but also that various policy options are on the table. Our institutions must be designed so that experts might be able to fulfill a certain kind of role whether it is as “thoughtful partisans” or “honest brokers” of policy options. If watching a no-holds-barred scientific fight does no good for the scientific enterprise in general, a no-holds-barred political fight does no good for the enterprise of democratic politics either.

What might these redesigned political institutions look like? We can start by looking at examples around the globe and at some of the experiments in deliberative democracy. But that is a topic for another post.

1. To be clear, Sarewitz is focused on environmental controversies but the argument can be extended to any technical-political controversy.

2. I try to avoid the word “innovation” but in this case, I really think it serves the purpose.

3. I borrowed this formulation from the sociologist Gil Eyal’s excellent book “The Crisis of Expertise.” As Eyal puts it: “The “scientization of politics” inadvertently causes the “politicization of science,” and the two processes reinforce and entangle one another in an unstable, crisis-prone mixture.”

4. In some cases, you do get actors who are arguing in bad faith. But even in those cases, such as the ones highlighted by historians Naomi Oreskes and Erik Conway in their book Merchants of Doubt, the scientists themselves tended to be sincere in their beliefs; it was often their funders who acted in bad faith.

5. Which begs the question: why did we end up in a place where we started to think that science could help us resolve our political disputes about policy? That needs a separate post.