MU Environmental Network News
March 2004
Vol. 10 No. 3

Editor - Jan Weaver
211 Lowry Hall, MU / Columbia MO 65211

Science and Environmental Policy - 7 Things Everyone Should Know

by Jan Weaver

1. It's not really about science. People may pretend that their decisions are based solely on scientific considerations, but there are other issues - cost, outrage, ethics, values and politics - that play a major role in the outcome.

2. There is no fixed definition or standard test for "Sound" science. People who use this term, or its opposite, "Junk" science, without specifying what is wrong with the methodology being used in a particular case are engaging in an ad hominem attack. All of the following methodologies are perfectly legitimate approaches to investigating scientific questions, though some may be more effective, ethical, useful, easy to implement or appropriate for particular disciplines - clinical trials, observational studies, case studies, surveys, epidemiological studies, lab experiments, meta-analysis, or computer modeling.

3. We cannot always do randomized, controlled, replicated experiments because they may not be ethical (like evaluating the effect of second hand smoke), the treatment may already have been applied (like the dead zone in the Gulf of Mexico), the system may not be replicable (like the climate system), or it may be prohibitively costly (like evaluating storm water impacts on streams). Therefore insisting that only these kinds of investigations should be used for shaping public policy is most likely a tactical maneuver intended to avoid having to change the status quo.

4. While there is no fixed definition of sound science there are warning signs of questionable studies. The presence of one or more of these should prompt further investigation - a vested interest in the outcome; overlooking or ignoring variables, gaps in data, or anomalous results; inadequate or biased samples; anecdotal evidence; correlation confused with cause; statements of certainty; lack of reference standards; lack of peer review; dismissal, without explanation or justification, of opposing or contradictory evidence

5. While anecdotal evidence and correlation may not - by themselves - be sufficient for drawing conclusions, they can be legitimate reasons for doing further studies.

6. It is unreasonable, from a scientific standpoint, to set arbitrary standards for certainty e.g. 95%, given the complex, systemic, non-linear and evolutionary nature of many environmental problems. Instead it is more appropriate to ask on a case by case basis how risky a policy is, and how sure you need to be. And from the standpoint of fairness, it is also legitimate to ask, who benefits and who loses if the risk materializes.

7. Two conflicting principles shape perceptions of risk in following a particular action
Precautionary Principle
reasonable - if there is some evidence of harm, move slowly
unreasonable - unless you are certain it is safe, don't do anything

Substantial Equivalence Principle
reasonable - if we have been doing similar things we should be able to assume similar outcomes
unreasonable - unless you are certain it is dangerous, full speed ahead

I borrowed heavily from these sources, and strongly recommend them if you want more background -

Peter Sandman Risk Communication

Frederick Kirschenmann What Constitutes Sound Science?

Council of State Governments State Official's Guide to Sound Science

Environmental Working Group Show Me the Science - Corporate Polluters and the "Junk Science" Strategy.