1) Vaccination protestors who refuse to have their children vaccinated for polio, measles, whooping cough, etc. There's a belief out there about a link between the trace mercury in older vaccines and the appearance of autism. I've never seen convincing evidence of causation from credible double blind studies, but it's a very important issue to a certain subgroup of parents. Many of the parents in the_child's class at school fall in this category.
Vaccination protest only works if the vast majority of other parents vaccinate their children regardless of the perceived risk from the vaccines. If too many parents skip the vaccines, the diseases return -- as whooping cough has begun to do -- with far more serious consequences to population of children, statistically, than the vaccines themselves.
In effect, parents who believe vaccines are dangerous are transferring the risk to the parents who don't believe it, and relying on those other parents to endanger their children to keep the unvaccinated children safe.
2) SUV drivers who have chosen their vehicles for the sake of accident survival and traffic safety. This is an oft-cited reason for purchasing an SUV. It's my understanding that the survival rate drops for people in accidents with an SUV involved where the second vehicle is an ordinary automobile, due to the mass differential, and the geometry of SUVs.
In effect, SUV drivers are transferring their risk to non-SUV drivers, raising their chances of survival by endangering the chances of the of the other driver.
I don't believe for a moment that vaccination protestors or SUV drivers are making choices with the intent of being deliberately destructive to others. I'm just noodling with the idea of how risk is transferred by these choices, rather than reduced or eliminated. Is it a social good that these risks get moved from one population to another, where the population experiencing the increased risk has no say in the matter?
The same logic applies to private gun ownership. Allow me to momentarily concede for the sake of argument that the most optimistic claims for the value of handguns in home defense and personal safety are valid. If Citizen A has a handgun and Citizen B does not have a handgun, and Criminal X is aware of these facts, Citizen B has assumed a meaningul portion of Citizen A's risk from predation by Criminal X. Does this imply, for example, that everyone should be armed to avoid this effect?
Again, no conclusion, just looking at a social phenomenon which interests me.