Are there situations where assuming the worst case scenario is a sound basis of policy?  To answer that question, we'd need to gauge the probability of the worst case happening as well as consider the potential risks and benefits of acting to prevent it from happening. 

First, the worst case scenario has to be a credible case scenario.  It surely isn’t enough to simply be able to imagine something truly awful – there has to be evidence. How much evidence is enough? And how does one extrapolate from current evidence to future events? There are, after all, lots of intervening and interacting variables that are bound to interfere with the smooth progression of the is to the will be.  How much uncertainty are we willing to tolerate in order to proclaim the worst case is plausible enough to act accordingly? 

These considerations wouldn’t matter so much is there were no costs to actions/policies based on worst case scenarios. If all actions and their effects were equal, then go with the worst case! Nothing to lose and everything to gain!

But is that the case? Are there no costs to government policies being based on the assumption that the worst case scenario will happen without government intervention?  Are there any downsides to going that route?