Are Brazilian crop data suspect too?
China's official agriculture commodity data have long been viewed with suspicion, misgivings borne out by huge revisions earlier this month to US government estimates for China's corn dynamics.
And doubts over Russian statistics, and that grain supplies were cut as low as had been reported by last year's drought, appear to have been borne out by the huge tonnages of wheat being reported at Black Sea ports ready for export. Agrimoney.com has heard trade estimates of 7m-8m tonnes of old crop grain alone needing to find a buyer.
Now Brazil's data is coming under the microscope after official statisticians pegged corn yields for the winter, second or so-called safrinha crop well above levels that farmers are reporting.
Farmers vs officials
Conab, Brazil's official crop bureau, has pegged yields of safrinha, or winter, corn in the top producing state of Mato Grosso as falling only 3.5% this year, leading to a crop of 7.0m tonnes.
However, Mato Grosso farmers are pegging crop losses as far bigger, of up to 50%, after a bet on a late onset of the dry season turned wrong.
Many Mato Grosso growers opted to plant safrinha corn late, well beyond the ideal sowing window, after a late soybean harvest kept fields tied up with standing crop far longer than had been expected.
Farmers were encouraged by revisions to dates for which they were eligible to claim on insurance policies.
'Definite difference of opinion'
"There is a definite difference of opinion concerning how the early onset of dry weather has impacted the safrinha corn crop in Mato Grosso," Michael Cordonnier at Soybean and Corn Advisor said.
"The official institutions are much more optimistic concerning corn yields than are the farmers in the state."
Dr Cordonnier has pegged Brazil's 2010-11 corn crop, of which safrinha accounts for some 40%, at 53m tonnes, 3.7m tonnes lower than the Conab estimate.
The USDA has a 55m-tonne figure.
Nor are these doubts over some Conab methodology a one-off.
In coffee, the bureau's has been consistently conservative – and perhaps overly so - on its coffee production data, to judge by comparisons with equivalents from USDA staff in the country.
"Historically, there has been a difference between our numbers and Conab's," an official US source told Agrimoney.com, saying the department's own in-country estimates were based on meetings with figures from throughout the industry.
Statistics from USDA attaches in Brazil "have been always consistent with the supply and demand spreadsheet - for example, ending stocks have never been negative in the end of the cycle", the source said.
"The same does not happen if you use Conab's number in a long-term data series."