It's possible for pressure to exist at a higher level. The editor simply stops asking reviews from people who reject papers above some threshold rate. Or at eve

It's possible for pressure to exist at a higher level. The editor simply stops asking reviews from people who reject papers above some threshold rate. Or at even a higher level, the journal replaces the editor who maintains a too low an acceptance rate. These pressures are harder to notice, without population level data.