Science has a credibility issue

Scott Adams, creator of the cartoon Dilbert, says that "...everything about science and nutrition." is Science's biggest fail. But I think it's deeper than that.
He sort of nails it here:

Science is an amazing thing. But it has a credibility issue that it earned. Should we fix the credibility situation by brainwashing skeptical citizens to believe in science despite its spotty track record, or is society’s current level of skepticism healthier than it looks? Maybe science is what needs to improve, not the citizens.

IMHO, he's right that science is what needs to improve. More specifically, the approach that scientists take to working on issues that matter deeply to us. I think all of his examples of how science said "Do this." followed a few years or decades later by "Ooops." are explained by scientists acting as "Stealth Issue Advocates". Nutrition. Climate Science. Feral Cat management. Grey Wolf reintroduction. In every case scientists have used their science to advance strong normative positions, and ultimately ended up politicizing and devaluing the science.

This is related to the recent Motherjones article that pointed out how different the opinons of scientists are from the general public on a range of issues. In almost every case, the question posed is one of risk tolerance, not facts. For example, 37% of the public agreed "It is safe to eat GM foods", while 88% of members of the AAAS thought GM foods were OK. The implication is that people lack the knowledge of the scientists, and that's why they worry more. But even college graduates are split 50-50. They kind of admit that a lack of knowledge might not be the whole story:

On some issues there are clearly factors beyond pure science, like ethics and politics, that influence opinions. For example, scientists show more support for nuclear power, but less support for fracking, than the public.

Clearly, if scientists are more supportive of nuclear power than the editorial board of MotherJones, something other than knowledge must be involved!

Where we get messed up, as scientists who want to be relevant and have impact, is when we mistake issues involving trade offs between competing risks for issues of fact and science. For example, on climate change, the fact is that increasing anthropogenic greenhouse gases increases energy input to the atmosphere. That has many, many effects on human interests, like cutting $20 Billion off of soybean yield over the last two decades. However, when it comes time to discuss what we should do about those impacts, we need to recognize that tradeoffs in human values dictate what will happen. Avoiding that future, imperfectly known sea level rise means I must give up my SUV*, my central A/C**, my cheap beef***. I don't even live on the coast!

Forget climate change for a minute. Think about feral cat management. Trap-Neuter-Release and variants are increasing in popularity as measures for cash-strapped municipalities to deal with increasing stray cat populations. Wildlife biologists and conservationists oppose these programs tooth and nail, because a spayed/neutered cat kills prey just like an intact cat does. The reason why the two sides disagree is that their goals differ. Catvocates want to minimize the number of cats euthanized at shelters. Wildlife biologists want to minimize the population of cats outdoors. No amount of science, none, will change that disagreement. Arguing that "science says we should trap-euthanize" is disingenuous. Science says no such thing. Science can predict future populations under different management regimes, but in every case trap-euthanize will euthanize more cats than TNR even if it reduces the population more quickly.

It is all about the values. The quicker we scientists learn that and practice it, the more effective we'll be at helping guide policy.

*I don't own an SUV. I do drive a big truck when hauling the Astragalus Field Lab.
**I keep the A/C at 78 in the summer, and we use way less electricity than the average LES customer.
***I only eat grass fed beef, which emits less methane, and that infrequently.