Scientific robustness - not certainty - is the key to good environmental regulation. Only when this is recognized will debates about the soundness of scientific advice or access to data be properly resolved.
Pity the US Environmental Protection Agency (EPA). For years, the agency has been attacked for the poor quality of the science on which some of its regulations have been based. Recently, it has been making strenuous efforts to correct such weaknesses, by, for example, setting up a board of scientific counsellors to help reorganize its peer review procedures, and quadrupling its spending on grants to external research teams. But the more EPA seeks to increase its scientific credibility, the more it has come under fire from industrial lobbyists and their allies in Congress.
Recent weeks, for example, have seen the resurrection of earlier proposals to establish a separate National Institute for the Environment (NIE, see page 612). Supporters argue that it is needed to ensure that environmental policy is based - and based only- on high quality science, and that that requires freedom from the intense political pressures which surround the activities of a powerful regulatory agency.
Nothing wrong in that, at least in principle. The problem lies in the questionable motivation and flawed understanding of the scientific process on which such campaigns are frequently based. The revival of interest in the environment institute has been largely prompted by industry's opposition to tough new restrictions proposed by the EPA on particulate emissions (see Nature 388, 5; 1997). Despite a large body of epidemiological evidence supporting these proposals on the basis of the thousands of premature deaths they would save, those balking at the costs of their implementation continue to insist that the evidence remains insufficient.
This is where the flawed understanding of science comes in. Thereis all the difference in the world between 'good' science, and 'certain' science. Science is continually evolving, and its conclusions are therefore never invalid just because they are provisional or uncertain. It may, for example, be impossible to distinguish a robust correlation between levels of exposure to small particulate matter and the resulting health damage. But that does not mean that there is no correlation, merely that the mechanisms involved remain obscure.
The same suspect motivation and misunderstanding underline current controversies about access to raw scientific data. There have been several recent moves in the US Congress to require federally funded scientists to release the raw data on which their conclusions are based. It is entirely appropriate that, in the appropriate conditions of commercial or personal privacy, access should be granted to those genuinely interested in testing the robustness of conclusions drawn from the data. But in practice this will work only if the difference between robustness and certainty is clearly acknowledged.
The United States does not have a monopoly on such controversies. Similar issues surround both last week's resignation of the chair of a French committee of inquiry set up to test claims of excessive childhood leukaemias around the nuclear reprocessing plant at La Hague, and the difficulties experienced by this inquiry in ensuring the robustness of the local medical data on which conclusions about possible leukaemia clusters have been based (see page 614). The former paid the price of excessive confidence in the safety of nuclear power. Hopefully, those concerned with the latter will carry out their task more open-mindedly.
Material presented on this home page constitutes opinion of the author.
Copyright © 1997 Steven J. Milloy. All rights reserved. Site developed and hosted by WestLake Solutions, Inc.