Challenger's disaster was the 25th shuttle flight.
Columbia's disaster was the 113th shuttle flight.
I'm pretty sure you could find large numbers of people who would happily climb on board a shuttle mission today, even knowing that history.
For policy purposes, killing people has a cost that can be estimated by the statistical value of a human life (especially if the people are volunteers with full knowledge of the risk.) The value of a human life is about $9 M (which comes from estimates of how much government spending is needed to save 1 life, for example by medical care, installing guard rails on roads, etc.). If there was a 2% risk of death of the seven crew on a flight, this would have added $1.26M to the expected cost of a launch. This was small compared to the actual cost of a launch.
Viewed another way: a $900 M (say) shuttle launch would be killing 100 statistical people each and every launch (in the sense that the money spent on the launch, if spent elsewhere, could avoid 100 deaths). If the results of the launch are worth that many statistical deaths, why not .14 more?