Canadians like to speed. A Transport Canada report on Canadian driving makes this clear:

Most Canadians will freely admit to exceeding the speed limit from time to time, but when it comes to just how fast they drive, they underestimate their true speed and delude themselves into thinking their behaviour really isn't that bad, says a new report by Transport Canada.

Based on a survey of 2,002 drivers conducted in 2005 and 12 focus groups held across the country, the report finds most Canadians consider speeding to be dangerous and believe it leads to an increased risk of collision, injury or death.

However, seven out of 10 Canadians admitted to doing it anyway, while an additional 11 per cent said they speed but only on highways, according to the executive summary of the forthcoming report, a copy of which was obtained by the Canadian Press.

When you think about it, there are three elements in play when someone is driving too fast:

1. the car's ability to speed
2. the road's design that tolerates speed
3. the driver's decision to speed

So how do we reduce speeding?

We can target the vehicle. Many models can drive at speeds much faster than is allowed by any jurisdiction in North America. The problem from a design point of view is that there are competing needs. We want the cars to fuel efficient, so we make them lighter. But when you combine that light weight with power requirements (minimum acceleration for safe merging, power needed to overcome drag and friction) you end up with a car that goes faster than you want.

There is a basic triangle at work here. You can define a car's weight (or at least adjust it significantly). You can define a car's acceleration and power. You can define a car's maximum speed. But you can only define two of those properties simultaneously -- the third is a function of the other two.

You could artificially force the maximum speed down with a governing device, which of course would be disabled by mechanics working for cash in a black market that would immediately spring up. In any case, people would argue correctly, what if the sudden burst of speed was needed to avoid an imminent collision?

So if we are stuck with cars that go too fast on a straightaway, then maybe we make fewer straightaways. This is the second point -- make the roads slower. More curves and turns to force drivers to slow down and maneuvre. Speed bumps that compel drivers to brake. Of course, the problem is that you can't retrofit the roads without spending billions, and we certainly can't bend them this way and that where they've already been built. New roads can be made this way, but not highways which need to handle truck traffic.

And that's where a lot of the speeding happens.

So what's left? The driver, of course.

But what else can you do? Increase the fines? More jail time? Take away their cars? All these approaches are being used, and they do work -- to a point. But like in most things in life, further increases of pressure result in less decrease in volume. It's a problem of diminishing returns.

This report acknowledges that by recommending a focused effort:

The report suggests the biggest problem group of drivers is the 30 per cent who speed more than average, and don't believe their speeding translates into a greater risk of an accident, injury or death.

They're labelled two ways, either as risk-takers or pragmatic speeders. Risk-takers enjoy the thrill of high speed and defying authority, while pragmatic speeders consciously drive aggressively when they believe they have reason to, such as being in a hurry.

The report suggests public education campaigns should focus on this hard-to-reach group because they pose a considerable danger to themselves and others, but seem oblivious to the risks until it's too late.

I have to say that I'm skeptical. If a person is too dense to know that his excessive speed is putting himself and others at risk, a series of TV commercials is not likely to change his mind.

The problem with these people is that unless they personally experience the problem with speeding, they always assume it's the other guy's speeding that is the problem. So maybe we need to have these people personally experience the benefit of not speeding on an ongoing basis.

Why not a special card that discounts a significant portion of the tax paid on gas? The card is passed to the pump like a loyalty card as part of the at-the-pump electronic transaction, or presented to the sales person during the checkout. I suppose for security, the card would have to have a PIN associated with it, otherwise people would try to take your card**. Anyway, the card is issued to drivers to use, but the card is disabled for a year or some other set period when a person is found guilty of a speeding violation (and perhaps certain other moving violations).

Here's what I like about the idea:

1. It is tied to money. People are reminded of the potential cost of speeding every time they fill up. And it is a positive reminder for those who are enjoying the discount.
2. The decision to disable the discount could applied more broadly for people guilty of lesser speeding offences, whereas now those speeders are often let off the hook since the current punishments seem a bit too severe.
3. It's good for the environment. People speed less and so use less gas.
4. The extra money collected from people forced to pay the fully taxed price for gas for past speeding offenses can be directed to funding for road work or for mass transit.

Could this idea work? I'm sure there are all sorts of reasons why it wouldn't. And yet it takes an entirely different run at the problem of punishing chronic speeders, and that suggests to there is room for such an approach to have a real effect.

** Maybe encode the discount on your driver's license? Built-in photo security to deter using other people's cards.