Well, the hiring committees see the entire prior rating distribution from each interviewer as well as the hires/no hires for each previous rating.

So the committees has a decent way of correcting for interviewers who use different scales.