The characteristic need not be common, just easily associated.
See the study here for rape and sexual abuse statistics: http://www.publicintegrity.org/investigations/campus_assault.... Even if we limit "rape" to situations involving violence and the threats of violence,[1] the study found about 1.3% of the college students questioned admitted to acts that met the definition of rape, and 1.1% admitted to acts that met the definition of sexual abuse of a child.
Now, for example, are even 1% of muslims terrorists? Almost certainly not, but the behavior is common enough that it is easily associated with the group, and readily supports many jokes relying on that association.
The relevant behavior does not need to be common, just common enough such that it is readily conjured up in the mind of the listener in response to the joke.
[1] If you include people who answered "yes" to the question of whether they had sex with someone who was "too intoxicated (on alcohol or drugs) to resist your sexual advances (e.g., removing their clothes)" the number skyrockets to 6.4%.
If you include people who answered "yes" to the question of whether they had sex with someone who was "too intoxicated (on alcohol or drugs) to resist your sexual advances (e.g., removing their clothes)" the number skyrockets to 6.4%.
Sadly, this study did not ask the question of what happens to that number if you reverse the genders. I have multiple male friends who meet the definition of rape victim -- but wouldn't dream of reporting it as such because they're in long term (and aside from the occasional misunderstanding, entirely consensual) relationships with their rapists.
See the study here for rape and sexual abuse statistics: http://www.publicintegrity.org/investigations/campus_assault.... Even if we limit "rape" to situations involving violence and the threats of violence,[1] the study found about 1.3% of the college students questioned admitted to acts that met the definition of rape, and 1.1% admitted to acts that met the definition of sexual abuse of a child.
Now, for example, are even 1% of muslims terrorists? Almost certainly not, but the behavior is common enough that it is easily associated with the group, and readily supports many jokes relying on that association.
The relevant behavior does not need to be common, just common enough such that it is readily conjured up in the mind of the listener in response to the joke.
[1] If you include people who answered "yes" to the question of whether they had sex with someone who was "too intoxicated (on alcohol or drugs) to resist your sexual advances (e.g., removing their clothes)" the number skyrockets to 6.4%.