Thursday, August 12, 2010

CP Blames Statcan for Bad CP Reporting

If one thing has been made clear during this whole census kerfuffle, it's that journalists aren't that great with numbers. Not to say that this wasn't clear already, but watching them report on something as number and method heavy as statistics gathering just puts a giant exclamation point on the fact. This story adds another exclamation point, followed, perhaps, by a "WTF?".

The headline suggests that Statcan is just rotten at its job. The first half of the article suggests that economists throughout the land are exasperated by Statcan's economic reporting. The damned bureaucrats don't know what the hell they're doing. Presumably, and I may have my partisan goggles on here, the author would like the reader to think that the government is right to give the agency a quick kick in the arse... even if the kick is poorly aimed and the government ends up on its behind.

Then you get halfway down and find out that actually it isn't really an issue. Real economists take individual job surveys "with a grain of salt" and pay attention to wider trends from multiple sources. This is followed by this hilarious line:

"That is what private sector economists also advise, but the caution often gets lost in the headlines about jobs created and lost."

And then it all becomes clear. It isn't that Statcan is doing a poor job. It's that the media doesn't know what they're doing and that an issue that isn't particularly new is now big news because of another completely unrelated story.

I get off my bus at Tunney's Pasture every day and walk past Statcan's office on my way home. Lately, if you listen very carefully, you can hear the cries of pain as mild-mannered statisticians pull their hair out.

1 comment:

  1. I saw this story and it's a pile of crap.

    While Stats Can releases the data, the private sector still develops their own forecast. If this issue has been prevalent for a few years now, the private sector modelling would already account for the discrepancy.

    Bottom line was that the private sector forecast was much rosier than the end result.

    If we've seen the same pattern for July numbers since 2004, that's hardly Stats Cans fault.