I think that a serious study has to be undertaken, with representative and random sampling. That would be the best way to proceed. We need statistics in order to truly assess what's happening. It would probably become apparent that in Quebec the situation is different because we already have a regulation that requires that retailers calibrate their pumps every two years. There truly has to be a serious study with serious random sampling as opposed to using data that may, for example, come from studies on consumer complaints. Obviously, that sampling is not representative. It is important to have random sampling.
On June 15th, 2010. See this statement in context.