Well... some researches do, actually, make the analogous argument (dubiously, IMHO) that people engage in riskier driving because of seatbelts (just like Paul is claiming for masks):

https://en.wikipedia.org/wiki/Risk_compensation#Seat_belts