First off, it was in the GLOBE too. Secondly, it went on for five years. Thirdly, I am familiar with a few other sectors of data that StatsCan collects and publishes and know these too are riddle with inaccuracies based both on the filing source material, and the understanding of the managers of the intricacies of the sectors they are covering off. I had a running battle for five years when I worked at the CRTC and we had no luck convincing StatsCan they were measuring the wrong things in the sound recording and film production sectors, not to mention making rediculous extrapolations from survey data about peoples media use habits. And it hasn't changed.