For the self-reported data that we have on victimization, we are not able to separate the information for on and off reserve because the sample sizes get really small. We have to remember that this is a survey of the general population. Basically, we're surveying everybody in Canada, and we're trying to produce estimates for everybody in Canada.
The aboriginal population, as you know, is about 4% of the Canadian population, so you could expect approximately 4% of the sample of that survey to be aboriginal people. The smaller the numbers get of records that we have, the less able we are to produce detailed information.