There is a strong case to be made, from an environmental perspective, in favour of owning multiple cars.

Wait, what?

Conventional wisdom is that the environmentally friendly thing to do is to not have a car, or if you depend on cars, to have as few as possible. In many cases (indeed, in any case where you can get by without a car), the conventional wisdom is right.

It turns out, though, that if two conditions are true- (1) you depend on cars and have no public transit options, and (2) you need a large vehicle often enough that renting a van on occasion doesn't make sense- the environmentally preferable option may be to own more vehicles, not fewer. Sharing the largest of these vehicles (something that's very difficult under current insurance rules, but could be encouraged with insurance reforms) would make the benefits easier to achieve.

Scenario

The analysis I'll present here applies to a specific set of circumstances:

• A family or group of neighbours lives in a suburb where public transit is limited or nonexistent.
• All adults depend on cars to get to and from work.
• Hauling (kids / dogs / trailers / supplies) to and from the (hockey rink / campground / school / store) is more than a small car can handle, so a 7+ seat van or SUV is required at least a couple of times a week.

If viable public transit exists, or if work is close enough to bike, walk or ski to the office, the situation changes. That said, there are a lot of Ontarians (and other North Americans) who are car-dependent and have no other viable transportation options.

Consider a middle-class family in an Ontario suburb. Both parents work in adjacent cities. Commuting by public transit would be a two-hour adventure spanning four routes on three different transit agencies- and repeat that in the evening. Commuting by car takes one hour to go 40 km each way. Very few people will consider transit to be a viable option here, choosing instead to commute by car.

Our example family's driving requirements are:

Commuting	$80 \frac{km}{day}\times 2 people \times 240 \frac{workdays}{year}$	38400 km/year
Groceries, shuttling kids, etc	$100 \frac{km}{week}\times 52 \frac{weeks}{year}$	5200 km/year
Trips with the whole family		4400 km/year
Hauling big stuff (lumber, furniture, etc)		2000 km/year
Total		50000 km/year

These figures are not at all unrealistic. I know many couples who, despite genuine attempts to limit their car use, put 25000 km/year on each of two cars.

Criteria

I'll evaluate the impact of all this driving using the following factors:

• Fuel consumption and cost, based on city fuel economy ratings provided by Natural Resources Canada.
• Damage to roads, which scales with the fourth power of per-axle weight. For simple comparative purposes, I'll state this in "Fiesta-km equivalent", i.e. by comparison to the net damage done to public infrastructure by driving our first example car, a 2012 Ford Fiesta, 1 km at its curb weight of 1170 kg.
• Carbon dioxide emissions, which is a simple stoichiometric conversion from fuel consumption.

Analysis

The need for a large vehicle on occasion means that, in many cases, such a family will own one small, efficient car and one great fat whale of a van or SUV. But we just saw that this family only needs the big vehicle for perhaps 6400 km/year of the 50000 km/year that they drive. Let's say they own these two vehicles:

• Car 1: Ford Fiesta (4 usable seats, 1170 kg, 6.8 L/100km = 157.76 g/km CO₂)
• Car 2: Dodge Grand Caravan (7 usable seats, 1950 kg, 12.2 L/100km = 283.04 g/km CO₂)

Then their annual fuel consumption will be:

• Fiesta: $25000 km \times 0.068 \frac{L}{km} = 1700 L$
• Caravan: $25000 km \times 0.122 \frac{L}{km} = 3050 L$
• Total 4750 L gasoline consumed (\$5700 at today's prices)
• Total damage to public roads: $25000 + 25000 \times (\frac{1950kg}{1170kg})^4= 192901$ Fiesta-km equivalent
• Total 11020 kg CO₂ released

What if we add a third car to the mix? Anything small and efficient will do; let's use another Fiesta for this example (although any conventional compact or small- to mid-size hybrid will give similar numbers). We'll keep the big thirsty beast- the van- in the driveway unless we're hauling more stuff and more people than the little compact cars can hold.

• Fiesta 1: $25000 km \times 0.068 \frac{L}{km} = 1700 L$
• Fiesta 2: $18600 km \times 0.068 \frac{L}{km} = 1265 L$
• Caravan: $6400 km \times 0.122 \frac{L}{km} = 781 L$
• Total 3746 L gasoline consumed (\$4495)
• Total damage to public roads: $25000 + 18600 + 6400\times (\frac{1950kg}{1170kg})^4= 92983$ Fiesta-km equivalent
• Total 8690 kg CO₂ released

By adding a third car to our two-car family, and adjusting our driving habits to avoid the big van when it's not needed, we've saved:

• 1004 litres (approx. $1200) per year in fuel.
• 2330 kg- or twice the extra car's weight- in annual CO₂ emissions.
• 52% reduction in the damage we cause to public roads.

The rule of thumb is: Always use the smallest, most efficient vehicle that is capable of doing the job. If only two vehicles are available and one of them's sized for your maximum conceivable load, you're bound to end up driving a thirsty whale of a truck on a trip for which an efficient two-seater would be perfectly adequate.

It is worth noting that, over the long term, this scheme does not appreciably change the economics of maintaining these cars. A properly maintained, periodically rust-proofed car will die of high mileage long before it dies of old age. Its repair and maintenance costs are, between 3 and 12 years of age, very close to a flat rate per kilometre. Driving two cars 10000 km each will cost about the same in maintenance as driving one car 20000 km.

The cost of buying cars is, on average, also no different for this three-car scenario than for a two-car scenario. A family with two cars might replace one of their cars every three years, i.e. they'll keep each car for six years. The same family in the three-car scenario described here will still buy a car every third year, but they'll keep them longer- nine years, on average, although the small high-mileage cars will be replaced more frequently than the low-mileage van. The rate at which money is spent buying new cars is unchanged, and the rate at which new cars are built is unchanged. They're just kept a bit longer before they're thrown away.

We can do better yet: It is easy to see that, if it were possible for a group of neighbours to each have their own small, extremely efficient car and share ownership of a few large serious-hauling vans or trucks, we could achieve similar improvements without the added hassle of owning an extra vehicle.

Bureaucratic Complications

Insurance, though, does not scale so nicely, at least in the arcane, convoluted insurance system used by Ontario and many American jurisdictions. Our current insurance system makes it very difficult to keep a large, rarely used vehicle sitting around. The cost of insuring the extra vehicle essentially wipes out the savings in fuel bills.

Our auto insurance system conflates the car and the driver. Each vehicle has a primary driver, whose liability insurance is in turn based on that vehicle. This works well when one person owns one car. With the addition of secondary and occasional driver clauses, it can cover a husband and wife on each others' cars, or children who occasionally drive their parents' cars in a two-car household.

When you try to share cars, though, things get very complicated very quickly. The three-car family described above would find, when their first child gets her licence, that each driver must be declared as primary on one of the vehicles. Listing a 17-year-old as a primary on anything sends the insurance premiums into the stratosphere, even if her driving habits are no different than they'd be as an occasional driver. You could keep the premiums down by allowing the teen to drive only one of the cars, thereby listing her as an occasional driver (much cheaper), but then you're in big trouble if there's an emergency and the only car the teen's insured on is parked at Dad's office.

Trying to share cars among neighbours is even more difficult. If my personal premiums are tied to claims involving a car I own, I'm going to be very reluctant to loan that car out to a neighbour. It may make economic sense for six families to share an Econoline and two Grand Caravans among them in addition to owning their own tiny, efficient commuter cars, but there is no easy way to arrange insurance for such a scheme.

Outfits such as Carsharing, Car2go, AutoShare and Zipcar have made considerable progress towards resolving these issues. It took a long time and a lot of legal fine-tuning by some very clever businesspeople, but they've reached a point where insurance coverage for city-wide car sharing fleets can be made economically viable. For folks who only need a car a few times a month, such services are convenient and are considerably cheaper than owning a vehicle. They don't, however, work very well for those who live outside dense urban areas, or for those who need a car on a frequent basis, and the insurance schemes they use are hard to scale down to neighbourhood co-op size.

A possible solution

Let's talk about splitting up car insurance.

Some parts of a policy relate primarily to the car- what kind it is, where it's based, how much it's driven, how often similar ones are stolen, and how much it costs to fix.

Other parts of a policy are mainly about the driver- how experienced he is, how many things he's hit recently, how much time he spends on what kind of roads, how often the cops flag him down, and how much money has historically been paid out in claims relating to drivers in the same demographic group.

These two aspects are typically conflated into a single policy that covers both the driver and the vehicle. Let's consider splitting them up into two components:

• A car policy that covers fire, theft and other things that relate mainly to the car itself.
• A driver policy that covers third-party liability and other things that are direct consequences of the driver's actions.

Now I can loan my van to my neighbour without fear that my car policy premiums will soar if he hits something- that would count against his driver policy. Four of us could pool our money to buy and share a truck; we'd each pay our own driver policies according to our own risk profiles, and we'd all chip in to pay for the truck's fire/theft/damage policy. Kids, when they get their licences, would get their own driver policies (possibly with restrictions on allowed vehicle types or other factors) instead of being tacked on to their parents' insurance.

The key to making it work would be to ensure that the total insurance cost for two people putting 50000 km on three cars is not appreciably higher than the cost for the same two people putting the same 50000 km on two cars. If that can be done, we can get a lot of single-occupant large vehicles off the road in favour of much smaller, much more efficient commuter cars.

The ability to insure a vehicle and a driver separately, with driver actions affecting only that driver's policy, would enable car sharing between families and neighbours on a much broader scale. This would remove the need for families to own- and therefore to commute in- their own large, fuel-guzzling vehicles. It would encourage the growth of co-operative car-sharing fleets on a neighbourhood scale, making it easier to always choose the most efficient vehicle for the task. Along the way, we'd see significant reductions in fuel use and emissions, as well as in road wear.

So, do we have any insurance execs or politicians out there who are interested in trying it?