Despite the warnings of risk communication specialists, members of the
technical community often urge that technological risks should be ''p
ut in context'' by comparisons against risks that are more familiar. L
ittle quantitative evidence is available on the actual behavioral cons
equences of such risk comparison efforts. In the present study, subjec
ts were presented with two types of information about a hazardous wast
e incinerator-a simplified statistical summary and a comparison of inc
inerator risks against the risks of smoking. Statistical information l
ed to a modest increase in the reported willingness to vote in favor o
f the incinerator in a community referendum, but the comparison agains
t cigarettes led to a slight decrease in support; the difference betwe
en the two messages is statistically significant (p < .001). In combin
ation with other results, this study's findings suggest that an implic
it assumption of risk comparisons is in error: Opposition to controver
sial technologies may have little to do with citizen's levels of infor
mation about technology, having more to do with citizen's levels of tr
ust in governmental and industrial actors.