Some of HRM’s richest neighbourhoods are being subsidized by some of our poorest ones, according to a new analysis by urban planners Tristan Cleveland and Paul Dec which looks at census data, road costs, and road length per capita. The analysis brings to light systemic inequities, and also shows that on average, Halifax has been growing less efficient in terms of managing our collective costs.

Of course, we’ve known for a while now that density is efficient and sprawl is expensive. But we’ve known it in the way many of us knew that smoking was bad for us long before we actually quit. We just seem to conveniently forget it at key moments, like when approving large-lot subdivisions in suburban communities.

Dec and Cleveland’s analysis might help us remember, because it puts the ‘sprawl costs’ argument in simple, more intuitive terms: “how much road do you want to pay for?”

All images from Tristan Cleveland and Paul Dec,

Cleveland and Dec counted the length of municipal roads in Halifax census districts (subtracting provincial roads and those with no residential addresses) and then used census data to calculate how much road per capita there was in each district. They found that suburbs like Glen Arbour (off Hammonds Plains Road, and home to its very own golf course) average 16 metres of road for each resident, well above the regional average of seven metres, and well above communities like Fairview which have three to four metres per resident.

It’s density spun down to a single dimension, road length, which for most of us is much more relatable to actual costs.

Dec and Cleveland then calculated the basic road costs for each district: snow clearing, annual maintenance and recapitalization costs, and compared them to how much each area was paying in municipal taxes. It turns out that Glen Arbour, despite being populated with expensive homes (which make for higher per capita property taxes) are an infrastructure burden to the city, contributing little to the city after basic road costs are covered. And communities like Fairview and Cole Harbour, it turns out, are not. Cleveland calls the effect cross-subsidization.

The road length analysis does not provide a full picture of municipal service costs, which get complicated to calculate because different parts of the city have different services. “That’s partially why the analysis looked at a pretty narrow set of costs,” says Cleveland. “Road maintenance and snow clearing have to happen anywhere you have a road.”

“If you look at how much road infrastructure there is per person in different areas, then you really can’t get away from the fact that there’s twice as much infrastructure we all have to pay for in one place versus another place,” says Cleveland. “The math at that point is as intuitive as two beers costing more than one.”

But there are limitations to the road length per capita analysis. First, it makes rural communities look incredibly expensive. So Dec and Cleveland have left out much of rural HRM from the analyses. “That’s not the point,” says Cleveland. “We’re not going to depopulate our rural areas. Yes, they cost money, but we’re going to have rural communities and that’s okay.”

What Cleveland and Dec are targeting is inefficient development in urban and suburban areas. They used the old Census Metropolitan Area for Halifax, “basically the commuter-shed, or everyone who drives to work in Halifax or inside that area,” says Cleveland.

And within that area, says Cleveland, “there really is no reason to build any community as inefficient as Kingswood and Glen Arbour.”

“We’re talking two- and three-acre lots, and that may be okay way out in a rural area, but for an urban setting where people are trying to drive or take transit to an urban core to work, the amount of inefficiencies and injustice it creates (because of cross-subsidization) is just ridiculous and unnecessary,” says Cleveland.

Cleveland and Dec looked at road length per capita in nine cities across Canada, and compared it to the typical measurement planners use, density. They found that as density decreases, at a certain point road length per capita increases exponentially. “Where efficiency falls off a cliff,” they call it.

“There’s this crucial threshold where the true egregious inefficiency starts,” says Cleveland. “We really think the city should just set a hard limit. There should be no development over six metres per person really, because anything higher than that takes us away from the goal to reduce the infrastructure burden per person.”

And what we really need to bring down our road length, says Cleveland, is infill development. “Every time we build on a road we already have we are decreasing the amount it costs each of us to run this city, while also bringing more people to our local business, to our transit–making everything we already have more successful while costing less.”

The team also recommends reforming the way taxes are charged to help reduce cross-subsidization, though Cleveland admits, fixing the problem on the supply side, with development limits and incentives, would be simpler and less controversial.

Bruce Fisher, HRM’s manager of fiscal and tax policy, cautions against tying taxes directly to services in order to fix these inequities. “This is not meant to be a service-based tax system,” writes Fisher in an email, although we do have some taxes tied directly to service, like our transit taxes. But generally, tying taxes to services transforms them into something closer to user fees, which if your goals are equity and efficiency, is not a great idea.

That’s not to say that our system is un-reformable. We could, for example, introduce progressive rates, like we have in our income tax system. With income tax, the richer you are, the higher the rate. With property tax, rich and poor pay the same rate.

Cleveland and Dec aren’t espousing a specific tax reform solution, but they are advocating a conversation.

“It’s worth asking, do we really have the right tax system? Is this helping anyone? Is this achieving any goal, if it creates these absurd situations of cross-subsidization while encouraging the kind of sprawl that we see?” asks Cleveland.

One weakness of focussing strictly on limiting road length for new development, says Cleveland, is that developers can reduce road length by creating communities with long curving roads that connect to nothing, making them highly un-walkable and difficult to service with transit. Cole Harbour, which looks peachy in terms of road length per capita, is an example of this.

So Cleveland and Dec are recommending that our planning policies couple an upper limit on road length per capita, with a minimum number of intersections per kilometre, to ensure that people in new neighbourhoods can actually get out of their houses without always getting in a car.

Road length per capita is an illuminating new indicator. So far, Dec and Cleveland have presented it at the Canadian Institute of Planners conference, and in Halifax to public and city staff.  Their next step is to incorporate jobs numbers into the mix when Stats Canada release those numbers in November.

“I really can’t wait to run the same analysis of road length per person and jobs,” says Cleveland. “That’s where you really start to express how efficient some places are compared to others. Whether people can work in a place has even more impact on whether transit is viable there and whether people can walk around. We’ll get a much more comprehensive picture of efficiency once we get those numbers.”

Join the Conversation


Only subscribers to the Halifax Examiner may comment on articles. We moderate all comments. Be respectful; whenever possible, provide links to credible documentary evidence to back up your factual claims. Please read our Commenting Policy.
  1. “With property tax, rich and poor pay the same rate.” Not exactly. Property taxes are effectively regressive, since lower income people pay a higher percentage of their income in property taxes. This is partly due to how property taxes are hidden in rental properties, and partly due to people who are low income but have valuable properties (which, if they sold, they’d have nowhere to live). So low income earners in dense neighbourhoods are doubly subsidizing the wealthy in sprawling areas.

  2. As someone who lives in one of the reasonably dense suburban areas, I like that this model demonstrates that reasonable density exists outside the centre core.

    But on the other hand, it doesn’t seem like much of a stretch to say say that urban is more dense than suburban is more dense than rural and that road length in an area is inversely proportional to density – what is a stretch is to arbitrarily decide that the rural lack of density is fine but suburban lack of density is not. Mixing data and value judgments is tricky.

    I also think that there needs to be some consideration given to the effect of time and infill on established neighbourhoods. It would be interesting to compare the denser suburban areas (that are 40-60+ years old) with what they looked like when they were newer and consider if there is the same potential for infilling in some of the less dense suburban regions over time

    Or we could just build hyper-dense capsule residences on the peninsula and require people to abandon everything else….wouldn’t that be efficient?

  3. There seems to be a typo in the first paragraph about how rich neighborhoods are subsidizing poor ones rather than the other way around.

    However it looks like the city is mostly positive, with only three red spots, one of which is McNab’s island, and one of which is that mostly undeveloped/used for military purposes area in between Dartmouth and Eastern Passage.

    People who live in places like Kingswood and Glen Arbor pay more federal and provincial tax, so just looking at the higher costs of road maintenance is not quite honest if you also analyze those places from the provincial and national perspective. Additionally, some people who live in desirable places like that might be genuinely talented professionals who might leave Halifax or Nova Scotia if they couldn’t live in a nice suburb 25 minutes from downtown – something you can’t do unless you’re incredibly rich in, say, Toronto.

    And yeah, it would be more efficient to put everyone in concrete megahives, I completely understand the technocratic argument, but some of us are misanthropes who might start shooting (or more realistically, drinking) to cope with living in close proximity to others. But hey, from a purely technocratic perspective, we might as well start euthanizing people who aren’t “efficient” and encouraging the breeding of the “more efficient”. The point is that as long as nobody gets a free lunch unless they’re too broke to buy one themselves it’s an okay but imperfect system.

    Also, road wear by vehicles increases by something silly like the fourth power of axle weight. So one use of a road by a tractor trailer might cause 1000 times as much damage as one use by a passenger car. The trucks are the real freeloaders, not people who live in Kingswood.

  4. Not much crime in the burbs, plenty of crime in inner city.
    Not much transit out in Kingswood etc.
    A credible analysis would consist of total HRM costs.

    1. Agreed. I’d love to know where revenues are originating in HRM and where they are being spent. The last time I asked city hall about this, I was told the data don’t exist. And Cleveland is right: we’re not going to depopulate our rural areas. But, by definition, they don’t belong inside “city” limits either.