To some extent this can be considered a success for Vector and the shift of our default from a skin optimised for editing to one optimised for reading. Of course if we want to increase editing levels we always have the option of defaulting new accounts to Monobook instead of Vector.

Typo correction and vandalism reversion are certainly both entries to editing, and it isn't just anti-vandalism where the opportunities have declined in recent years. Typos are getting harder to find, especially in stable widely read articles. Yes you can find plenty of typos by checking new pages and recent changes, but I doubt our 5 edits a month editors are going to internal maintenance pages like that. I suspect they are readers who fix things they come across. It would be interesting to survey a sample of them I suspect we'd find many who are reading Wikipedia just as much as they used to, but if they only edit when they spot a mistake then of course they will now be editing less frequently. And of course none of that is actually bad, any more than is the loss of large numbers of vandals who used to get into the 5 edits a month band for at least the month in which they did their spree and were blocked.. The difficulty of getting precise measurements of "community health" makes it a fascinating topic, and with many known factors altering edit levels in sometimes poorly understood ways we need to be wary of oversimplifications. No-one really knows what would have happened if the many edit filters installed in the last four years had instead been coded as anti vandalism bots, clearly our edit count would now be much higher, but whether it would currently be higher or lower than in 2009 when the edit filters were introduced is unknown. Nor should we fret that we shifted so much of our anti-vandalism work from very quick reversion to not accepting edits. However it isn't sensible to benchmark community health against past edit levels, we should really be comparing community activity against readership levels. If we do that there is a disconnect between our readership which for years has grown faster than the internet and our community which is broadly stable. To some extent this can be considered a success for Vector and the shift of our default from a skin optimised for editing to one optimised for reading. Of course if we want to increase editing levels we always have the option of defaulting new accounts to Monobook instead of Vector. My suspicion is also that the rise of the mobile device, especially amongst the young, is turning us from an interactive medium into more of a broadcast one. It is also likely to be contributing to the greying of the pedia. I am trying to list the major known and probable causes of changes of the fall in the raw editing levels in a page on wiki< https://en.wikipedia.org/wiki/User:WereSpielChequers/Going_off_the_boil%3F feedback welcome.