Some ambiguity and a touch of hyperbole in my post about targeted infill led some commenters to think that I was suggesting all of Karori and Churton Park were "distant hills" with "two buses a day if you're lucky". That's not quite what I meant to say, but I stand by my assertion that the outskirts of these suburbs (and many others) have poor public transport service, and thus poor public transport use, meaning that infill in those places would just result in more people driving to work.
To test that, here's a map showing the proportion of people in the northern suburbs who took a car to work, based on 2006 Census data. Shades of red show where more than half took a car, while in meshblocks with shades of blue, less than half took the car. Central Johnsonville is at the bottom left, with Churton Park at the top.
It's clear that the further you get from the transport hub of Johnsonville, the more likely you are to drive. It's not rocket science, but it shows that the mere presence of a couple of bus routes is not enough. Churton Park does indeed have a decent bus service at peak times, but many parts of it are a long walk from the bus stops and there's not a lot of choice. Central Johnsonville, by contrast, not only has more public transport choice (and I'd suggest that the option of rail as well as bus is an important factor) but also a certain amount of employment, so that some people can walk to work. It thus seems eminently sensible to do what the council is doing and limit the amount of ad hoc infill, while targeting residential intensification where there is a proper transport corridor and plenty of shops and services: places like Johnsonville town centre for example.
The above map shows just a small part of the city, and could give the impression that driving to work is the norm in most parts of town. Looking at all of Wellington City, by comparison, you get a very different picture:
Blue dominates in most parts of the city, except the northern suburbs and a few hillside and coastal fringes. As today's Dominion Post article about the rebuilding of the trolley buses says, "Wellingtonians [catch] the bus more than residents of any other Australasian city", and if you add in walking, cycling and trains, it's clear that most of Wellington (except wherever John Morrison lives) has a different attitude to cars from the rest of the country.
If you look closely (click the map for a much larger version) you'll see that Karori has mostly 30-50% car use, but with definite red patches, especially once you get away from the main drag. That's another example of how infill should be targeted at quite a fine scale. Increasing density by a moderate amount on the fringes wouldn't be enough to justify many extra bus services, but increasing density near existing transport corridors would make much more sense.
This isn't a proper analysis, of course: there's an element of "chicken and egg" about the whole thing, and it's hard to tell whether low public transport use is due to poor services or vice versa. The council will presumably be carrying out much more detailed GIS analysis (including transport catchments, existing density and the physical capacity for more building) and more qualitative analysis of heritage, character and attitudes before finalising the areas of "stability", "limited infill" and "change". Purely from this map, though, it's clear that the urban spine (Johnsonville, CBD, Newtown, Kilbirnie) is an obvious place to start, especially when you include the hospital and airport as major destinations. Parts of Karori, Wilton, Brooklyn Island Bay, Hataitai and Miramar also look feasible, and moderate increases in density there could justify even better public transport services, more local facilities, and thus an even greater proportion of people making sustainable transport choices.