Alt title: Why does Stats Can Suck So bad?
Let my qualify. Stats Can, on the collections end, is a very robust and professional organization. However on the data delivery end it is an onerous, counter-intuitive, expensive service with a Byzantine logic to the organization of its data. Really, I had forgotten how crappy it was until I had to explain to an undergrad how to go grab some time series data on real hourly wage rates going back 76. Keep in mind the student was accessing the data through CANSIM II apparently the research friendly portal to Stats Can data. I would hate to see what they do when they are not trying to be user friendly.
Stats Can seems to have the attitude that their data should presented in the most convoluted least user friendly way. Presumably to keep the non-experts out. And what is with the shitty state of the provincial time series data and the general refusal to make consistent series that go back to at least the 1960s. Either Stats Can is cashed starved or it is run by the equivalent of old-school linux geeks that have no time for pre-compiled drivers.
There is a reason why so many researchers who work with time series data choose to work on the US. First you do not have to be a member of Academic institution or a wealthy Think Tank to get access to the data: it is on-line and its free. Second you do not have to know the exact title of the series you are looking for. Indeed on some of the official US sites the data is thematically organized in easy to interpret categories. Third many of their official sites provide internationally comparative data sets. Next time I am sending my students to the American data.