These specific people would, yes. But how many more people would risk driving drunk if they knew there were no sobriety checkpoints?
Why do you feel a study is needed in order to form an opinion on this? What logical arguments would you make against sobriety checkpoints?
In any case, turns out that at least one study has been performed. This was just the first google search hit. There's probably more out there.
[1] http://www.cdc.gov/motorvehiclesafety/impaired_driving/check...
http://www.scpr.org/news/2014/01/06/41411/california-drunk-d...
BTW my "common sense" says that sobriety checkpoints DO reduce drunk driving and fatal accidents. I'm not any kind of DUI checkpoint denier.
It works here, a lot of people are preferring a cab over the car to go out to places they know they will drink.