Ahead of the next World Agricultural Supply and Demand Estimates (WASDE) report on Nov. 10, USDA officials are stepping forward to clear the air.
Processing Error in Report Was Computer Weighting Issue
In their first on-camera interview since admitting a processing error in the September data, officials clarified that the mistake was a computer weighting issue.
Speaking to AgriTalk PM host Michelle Rook, USDA Under Secretary for Research, Education and Economics, Dr. Scott Hutchins and National Agricultural Statistics Service (NASS) Methodology Division Director Lance Honig further explained the number of objective yield plot samples taken did not have an impact on the error that primarily affected corn yield estimates in Kansas and Nebraska.
USDA Revises Kansas, Nebraska and National Corn Yield
The net impact of the correction was a 0.3-bushel-per-acre increase to national corn yield to 178.8 bushels in September. This also included a 5-bushel increase for Nebraska corn yield and a 3-bushel per acre cut in Kansas.
However, as USDA has maintained, the October yield forecasts were fully validated, and were not impacted by the September processing error. This the October national corn yield remained unchanged at a record 181.2 bushels per acre nationally.
USDA Pushed to Provide Transparency
The exclusive explanation comes after Farm Journal’s tenacious pursuit of answers on behalf of producers. When NASS initially announced the data correction on Oct. 23, the agency cited a “processing error” but provided no details on what the error was or how it happened.
AgriTalk AM host Chip Flory heavily scrutinized the vague admission, relentlessly pushing the agency to explain the failure in order to restore market confidence before the November numbers are released.
“While it’s a good thing NASS has acknowledged the error and issued a correction, not knowing what the error was, how it happened and how NASS plans to not let it happen again raises many more questions,” Flory argued prior to the interview. “We will continue to ask questions that need answers in order to be confident in the process that provides critical information and data to the markets that has a direct effect on farmers’ businesses.”
Answering that call for transparency, Honig explained the issue stemmed from a file used to process September corn objective yield data. In Kansas and Nebraska, where the irrigated portion of the crop is highly significant, NASS collects samples independently but then must weight them to accurately represent the distribution of irrigated and non-irrigated practices.
“Obviously the error in the file means wrong weights were applied. And so once you do that, that means that you’re impacting the combined plant populations, the ear counts, ultimately those yield indications,” Honig explained.
Report Error Not Tied to USDA Personnel Cuts
Crucially, Honig noted the raw data itself was sound, easing some industry concerns about staffing shortages or enumerator failures in the field.
“I will say specifically though that the individual sample data that was collected by our staff in each of those fields was not impacted. It was all about how those results got weighted together to drive those totals that feed into those official yield and production forecasts that we published,” Honig said. “It had nothing to do with the number of samples. It was all how they were weighted together. And so that file, that incorrect acreage weights in it, was the cause of the problem.”
The error was initially flagged by observant data users—including Flory and other market analysts—who noticed glaring anomalies in the September report. Specifically, analysts questioned how September data could suggest 90% of Nebraska fields sampled were non-irrigated, only to shift to a more typical 55% irrigated balance in October.
“So one of the things that we do proactively, Michelle, is that we listen to the user community quite a bit,” Hutchins said. “Through all of our listening posts, there were several users who’ve paid very close attention, which we love. It noticed some anomalies that appeared to not fit... We initially looked at the October report to make sure that there were no issues or anomalies there and out of an abundance of caution, we also went back to September and that’s when we identified the anomaly that turned out to be the error.”
USDA Works to Rebuild Trust With Farmers
With margins already tight and farmers frustrated by the market volatility triggered by the October report, USDA is now focused on damage control and data modernization. Honig noted the agency is exploring new technologies, including artificial intelligence, to ensure validation steps “are as thorough as possible and can happen as efficiently as possible.”
For Hutchins, the path forward relies heavily on communication and the continued participation of farmers in NASS surveys.
“Well, Michelle, one of the ways that we rebuild trust, or build trust, is by being transparent. And as soon as NASS identified this opportunity and this error, they corrected it, they rechecked it,” Hutchins said. “Trust is not something we can ask for. We know we have to earn it. And we are working to do that through ensuring that we get the very best data, the best resources.”
