|►◄ Reverse Zone|
Complete List of Posts
Tue, 04 Oct 2005
This post will eventually become a longer article and who knows maybe a book some day.
I have a vision of a city designed to minize GHG emissions, whose profile is exactly the opposite of how cities are built now, and even very different from "Smart Growth" development as it is generally understood. Along the way, it clashes a bit with common sense, economics, and Euclidean geometry.
A year ago, I wrote a paper on the the density distribution characteristics of a municipal area that would minimize the total use of land. After a bit of math, and driven mostly by the behaviour of households with children, it turns out that a uniform density everywhere minimizes the total land use. For most cities, the "ideal" uniform density is one where the demographic statistics of each neighbourhood roughly match those of the entire city.
The density distribution characteristics of a municipal area that minimizes the total emissions of greenhouse gases is less uniform.
It is well known that on average people who live in higher density and closer to downtown use their cars less. This locational relationship to the amount of driving affects high-income households more than low income. Income, household size, vehicle ownership, distance from downtown are all highly correlated with density so it's hard to tell to what degree increasing density while keeping other variables constant will have an effect.
The distance from downtown has a greater than linear effect on vehicle kilometres travelled. If work is further away, you have further to drive. You are probably also part of a self-selected group who values the freedom of driving and whose finances allows the expense of driving.
If you use transit, then your distance from downtown does not affect GHG emissions significantly. The bus or train travels from one end of the route to the other no matter where people get on. Looking strictly at distance, if a household is very likely to use transit, they can live near downtown or near the end of the line with no significant effect on GHG emissions.
The number of kilometres driven by a household also depends heavily on the number of adults, and also on the number of children, the number and size of incomes, and the number of cars per household.
The type of neighbourhood you live in also has a great influence on how many kilometres you drive. Quite apart from driving shorter distances, people in traditional older neighbourhoods drive fewer trips. People in "new urbanist" neighbourhoods with more of a grid system also tend to use transit more.
Part of this effect has to do with having a walkable neighbourhood (see this site) When you live close to where you want to go and the walk is pleasant, you are more likely to walk there. This includes walking to the transit stop. I will not go into what makes a transit stop attractive, there has been a lot of research about this, but distance and the quality of the route to the transit station contribute to transit ridership.
There is a certain minimum density required around transit stations to make the route economically sustainable. However many planning authorities have extended this principle to maximizing density near transit stations. Unfortunately, this means that the people who you most want to take transit - the larger, richer households - are going to be living further away from transit, in the lower-density areas that they always seem to prefer. This is something which is difficult to communicate to the average urban or regional planner: density is a factor in sprawl and GHGs, but only low to medium density. Apartment dwellers, living in high density, are not part of the problem and they are not part of the solution.
So if the goal is to minimize the amount of GHG emissions, the shape of the city you need is not what you'd expect. You have to put all the people that are highly likely to take transit, including the poor and the old, the people most likely to live in high density such as apartment buildings, at the end of the transit line. Those that are likely to use a car no matter what, the rich, families with children and two incomes, should go near the centre where they will drive shorter distances. To maximize the probability that the high-transportation groups will walk or use transit, the neighbourhoods where they live have to be walkable. These groups tend to shun high density. The areas that are particularly convenient for transit should be attractive and relatively low density.
It seems heartless, but the poor and the elderly will use transit even if it requires more inconvenient connections or a longer walk. They would be happier if transit were convenient, but a rich white family with children and two incomes will burn five to ten times as much fuel as an older, poorer single-adult household if they choose to use a car. A small empty-nest household will drive 10 km a day on average. A high-income household with older children will drive nearly 70 km a day.
So in terms of residential density, what is required is relatively low-density housing near the central business district, with large units. Near transit stations further out, you need convenient village centres with a walkable grid and all the required amenities for daily life, including local stores, schools, community centres, and so on clustered around the transit station and again surrounded with relatively low to medium density housing around the station. The higher-density apartment buildings are on feeder lines leading to the suburban transit stations.
This is the density distribution that minimizes greenhouse gas emissions. It is the opposite of how cities are currently being built. The density is not maximized near the centre, as "highest-and-best use" dogma demands, nor is it maximized around transit stations as the most popular misconception of transit-oriented development is now saying.
Employment, on the other hand, needs to be highly concentrated and on main transit lines. That means segregation of employment from residential land use. Employment use should be as dense as the market will bear, while residential use should remain relatively low-density. This is the opposite of the "mixing of uses" that is preached in the smart growth arenas, although most new urbanism advocates do tend to keep high-density employment away from residential uses. I think it was Andrés Duany who said people want to work close to where they live but they don't want to live close to where they work.
This type of density distribution may sound odd, even impossible, but it describes several European cities, including Paris. The centre of the city has severe zoning that keeps the height down. Further out are the traditional village cores turned into suburbs, clustered around the train station. Between them are the high-rises with the poor, the immigrants, and others who can't afford the more traditional lifestyle of the city or the older villages.
There is plenty wrong with this model for reasons other than greenhouse gas emissions. First, although creating ghettoes of old and poor people in less convenient neighbourhoods minimizes GHG emissions, it is even more reprehensible a social policy than are the ghettoes they are currently being confined to. Secondly, Euclidean geometry says there is less room near the centre to fit all the single-family houses. It helps a bit that families of equal income tend to have smaller lots if they are near the centre.
Finally, economics in urban centres has a very odd effect. Land near the centre is desirable because it requires less driving to work. So desirable that the price goes up beyond what is realistic for building the larger units that households with several incomes require. The only economically viable housing form is the high-rise. Ironically enough, the people who end up living there are not the ones for whom the location was most useful. The ones who wanted to drive a shorter distance to work can't afford to live there. Meanwhile those who have little money to spend on housing end up living on the most expensive land, while the cheaper land is being used for those with more money. I thought the invisible hand was supposed to distribute goods and services in a way that maximizes utility. This is what zoning is for; to intervene where economics alone does not solve problems.
For general figures on auto use, see the table in