Road Length Per Resident

It can be remarkably difficult to convince folks density lowers costs. Here’s a different approach.

  • No one wants to be on the hook for a bigger chunk of road to pay for. So if we calculate how much road there is per person in different areas of the city, we can figure out which ones are creating a problem we’d all like to avoid.


  • So we measured the length of road per person for every census tract (i.e. neighbourhood) in 10 cities across Canada. Turns out, as density drops below a certain point, the road length per person starts to go up exponentially. We see a very similar pattern in every city.


  • Startlingly, we found that for Halifax, the more an area costs to service, the less tax revenue it generates. The more spread out homes are, the less potential there is for assets to be located nearby, and so the same factors that drive up cost. drive down land value—reducing property tax revenue. The more development costs cities, therefore, the less they chip in. That’s a bad deal.


  • When we take the total tax revenue of Halifax census tracts minus the total cost of road maintenance, we get a picture of how much each is chipping in above and beyond their road costs. We find the higher the road length per person, the less census tracts contribute.


  • Outside the core, residential tax revenue only adds up to enough to pay for road maintenance, meaning these areas don’t pitch in for the rest of the costs of the city. Many of the poorest neighbourhoods in the centre are, perversely, subsidizing high-income neighbourhoods on the periphery.


  • Road length per resident turns out to also be a powerful indicator for understanding healthy transportation. Once there are 7 meters of road or more between residents, almost no one walks or bikes, for any of the 10 cities we looked at.


  • We see a similar threshold for transit. If you want to make sure no one walks or takes transit, all you need is to make sure there are 7 meters of road per person or more.


  • Longer roads per person also increases the number of cars per household, greatly adding to the effective cost of houses.


  • We should use road length per resident to set regional targets, because all taxpayers can get behind having lower costs to pay for. This clear indicator can also be used as a useful tool for setting minimum standards for new development, to avoid the most egregious of inefficient growth. Finally, when combined with an analysis of tax revenue, it can help inform efforts to ensure some neighbourhoods (especially low-income neighbourhoods) are not subsidizing others.