The problem with over-reducing risk is that by eliminating a small risk (dammit, I hate the jargon. I'd much rather use risk as the combination of severity and probability of undesired effects) is that if the probability of damage is slight and the severity is great, we often wind
up banning something that would otherwise benefit the vast majority of the people involved. Take VIOXX, f'rinstance. For many pain sufferers it was about the only thing that worked. I know several well educated, intelligent folks that would be willing to take the (unlikely but severe) risk of taking it for the benefit of dramatically reducing pain.
Or consider the banning of hexachlorophene in hospitals--because a number of nurses failed to use it properly (failure to rinse it off) it's no longer used in hospitals. As a direct result, bacterial infection rates soared.

Identifying risks is a great and noble practice. Banning substances that may cause suh risks is often a serious mistake.