(Resource News International) — Statistics Canada, in its release of crop production survey results last Friday, surprised the industry by increasing its 2007-08 production estimate for canola — not because of extra supply, but rather to avoid a negative ending stocks forecast, according to an official with the government agency.
Statistics Canada, on Aug. 22, pegged 2008-09 canola output at 10.375 million tonnes, which compares with pre-report expectations that ranged from 10.1 million to 11.2 million tonnes. Canola output in 2007-08, meanwhile, was revised upwards to 9.528 million tonnes from 8.751 million.
The 777,000-tonne upward revision to 2007-08 production was first linked by the industry to higher than expected old-crop supplies.
“That was not necessarily the case,” according to David
Burroughs, head of the crop reporting unit for Statistics Canada.
He said the government agency, when it does its survey for July 31 production, also collects data for its supply disposition balance sheet for grains that are produced in 2007-08. That supply/demand data is supplied to private paying individuals.
“In gathering the data for canola, I recognized right off that there had not been enough production in 2007, as data from the Canadian Grain Commission showed that deliveries to end-users was greater than the supply base,” he said. “That had to mean we had underestimated 2007-08 canola production.”
Burroughs said it was more prudent to raise the 2007-08 production estimate for canola in the Aug. 22 production survey update than it was to raise the figures in the Sept. 10 grain stocks in all positions survey report.
As for why Statistics Canada had such a low canola production estimate for 2007-08, Burroughs linked the problem to old survey framework data that was based on information provided by the 2001 Census of Agriculture.
“There is sometimes a fair amount of deterioration in the framework data over the five-year period until the next Census of Agriculture is available,” Burroughs said.
Census of Agriculture data for 2006 was released in December, which helped to update the framework in which the data is collected.
Burroughs indicated that there is always the possibility that when the survey is being conducted, that farmers underestimate the size of their crops, et cetera.
“However, you always have the possibility that some farmers will say the crop is lower, and some will say it is higher,” he said. “Generally, that in itself has a way of balancing out.”
Burroughs said there was no way Statistics Canada can second-guess the information producers provide.
“We are asking farmers to give us their best guess and we have to pretty much accept those responses,” he said.
Burroughs said another factor that may have resulted in the upward revision, was that the survey framework was working with a crop land base that was smaller than in actuality.
“There is a sense that the Census of Agriculture data for 2006 may be working with a crop land base that is smaller than what actually exists,” Burroughs said. The survey for the 2006 Census of Agriculture may have been conducted too soon, as a year later, we are beginning to see that it is quite possible that a lot of crop land may have been missed.
Private industry officials have suggested the numbers for wheat and barley will also be raised in the Sept. 10 stocks in all positions report by the government agency.
Burroughs would not rule it out, but was unable to provide any details as the agency was still finalizing the data for the Sept. 10 release.