After years of study and millions (or tens of millions) invested, a report shows that we cannot yet tell whether the plastic molecule BPA (bisphenol-A) , to which we are all exposed, in bottles and can linings, is safe. BPA apparently acts like estrogen in experimental animals and cell culture. But is it harmful to humans?
Some scientists, whose job it is to try to look out for public safety, are concerned and say that we should limit exposure, and they want to prove that there's good reason for that. Some Republicans, whose job it is to look out for private pocketbooks, say that this is like yet another Commie plot, and they don't need any evidence, thanks very much (and, by the way, perhaps nowadays it's an Evolutionist or Islamic plot).
We don't know the answer. Sometimes in issues like these, the safe and sane thing to do is ban the questioned substance and just do without what usually minor loss it would lead to (except in the pockets of a few in industry who make whatever-it-is). In this case, however, the issues are less clear. Food packaging is important or even central to our general health and well-being. We have to have at least some food transported from large distances, and hence preserved in some way to make the trip without perishing. It's not all junk food that's involved. Most of us live in cities and know more about growing tired than growing carrots. Indeed, without our distribution system, would we live any longer in the wild than an escaped laboratory mouse?
So these issues go beyond objection to private self-interest by industry. But if science is so powerful, why can't we know the truth?
One response is that people, exposures and people's responses to exposures are highly variable. And, it's difficult to measure exposure, or even to know how to measure it, and people's responses can be so variable that it it's almost impossible to come to a conclusion about cause and effect. The evidence is weak because the net effect, the average effect, is small. However, for some people or some exposures, the effect could be huge. Maybe they have an unusual combination of exposures, or an unusually susceptible (but rare) genotype. It's very hard to know.
Alternatively, the effect could be similar to all of us but very, very small. In that case, what we would need to know is whether any alternative to using BPA would be without at least as many negative consequences. For example, if some people would have lessened nutrition and hence higher disease rates, if they were not able to obtain inexpensive foods preserved in part by the use of BPA. Even if a small number were helped, is that number larger than those who might be negatively affected?
Maybe when the risks appear to be very small, the best policy is not to worry. The risk may change year to year with changes in exposures or other factors. It may not be evaluable. On the other hand, what if there is some cumulative or long-delayed effect (a concern with genetically modified plants and the evolution of resistance, for example). Then we should try to determine this now, and stop the exposure.
We are constantly faced with such issues in evaluating causation. We have no answers. Unfortunately, nobody else does, either.