Yup, went up when they did an all postal ballot in some regions, but then steadied off.

But, there's some evidence that the bigger swings are voting to not voting, and vice versa. So while turnout is the same, different people vote each time.

I don't think enough studies have been done though, certainyl none I've got access to, might ask a friend who still has JANET access to do some digging, it's close to his PhD field anyway. But not immediately.