As evidence that this can occur, consider the US - inflation was known to be overstated by 1.3%/year before 1996.
https://en.wikipedia.org/wiki/Boskin_Commission
We probably do better by now, but there are still fairly well known upward biases in our inflation calculations. For example, consider the lack of hedonic adjustments to medical care in CPI. If an laparoscopic surgery today costs more than living in pain in 1970, the BLS counts the cost increase in "medical care" as inflation.
What's today's equivalent of the $800 20 inch CRT television set of 1990? A 40 inch LCD tv? With ($700) or without ($500) 3D? Or maybe a television and a $600 tablet?
So, it looks like a disagreement about the way to calculate inflation, with the challenger shouting hard, as he should, since it is the only way to make an impression.
If the challenger's arguments have merit, I guess the statitics bureau will slowly react by incorporating them into their computations. 'Slowly', IMO, is a virtue here. These computation methods should be designed by committee.
I don't know if inflation calculations were the problem in Canada, I'm just pointing out that it's one possible way.
Incidentally, there is a completely non-subjective way to calculate inflation: choose a fixed basket of goods. I.e., look at what someone in 1970 actually purchased, and then look today at what those goods/services [1] cost today. The only problem with this is that to calculate inflation from 1980 to today, you need to pick a 1980's basket rather than a 1970's basket, i.e. CPI stops being a single quantity you can push forward and backward in time.
[1] Also, don't choose categories of goods like health care - choose actual goods like appendix removal or aspirin.
That 'choose' part is subjective. http://www.bls.gov/cpi/ (stress added):
"The Consumer Price Indexes (CPI) program produces monthly data on changes in the prices paid by urban consumers for _a_ representative basket of goods and services"
Your 'or at the very least a 20" TV' acknowledges that, too. How do you objectively choose between the two?
Similarly, "like appendix removal or aspirin": who determines that "aspirin" is the actual good, and not "pain killing" or "prevention of strokes"?
> Statistics Canada issued a response... saying the
> agency follows international guidelines and practices.
> Given that Canadian productivity is frequently compared
> with that of the U.S., the methodology behind the
> estimates for Canada “should be comparable to the
> largest extent possible to that used by the U.S.
> Bureau of Labour Statistics,” said Wulong Gu of
> Statscan’s economic analysis division.
The difference between "shrinking since 1977" and "growing a percent a year since 1977" is so huge (at least 50% productivity) that I cannot believe that this is true.
Are people working fewer hours than the statistics bureau is aware of? If so, what do these people do in their spare time that does not show up as a "hey, that's funny" at the statistics bureau?
Or are there scores of 'died, but still reported as working' workers?
Alternatively, if the productivity is in "output per unit of capital": how do they hide using fewer resources from the government? Stockpile oil amd machinery? Or do they produce more, but sell on the black market?
For at least one of these, the estimates that the statistics bureau makes about them would, by now, have to be off by at least 20% or so. I think they would have noticed that.
On the other hand, I can easily see them as underreporting productivity growth by a few tenths of a percent each quarter, especially when reporting shortly after each quarter ended.