USDA Explains Corn Forecast Error After Review of Kansas and Nebraska Data

The U.S. Department of Agriculture has traced an error in its September 2026 corn forecasts to incorrect acreage weights used to combine irrigated and non-irrigated field observations in Kansas and Nebraska. The mistake changed yield estimates for both states and the nation, but USDA says the more recent October forecasts were checked and remain unchanged.

In an explanation issued Saturday, October 24, the National Agricultural Statistics Service (NASS) put Kansas’s corrected September corn yield at 119 bushels per acre, three bushels below the figure previously reported. Nebraska’s estimate rose five bushels to 182, and the U.S. average increased 0.3 bushel to 178.8. Those are corrections to earlier forecasts, not newly reported harvest results.

The clarification follows an October 23 notice announcing that September corn data had been corrected. Questions raised after the October 9 Crop Production report prompted NASS to review how it produced estimates for both months. The agency’s October 24 explanation identifies the problem as a September processing error and describes additional checks it plans to make before future data releases.

How the Kansas and Nebraska weighting error occurred

NASS said the calculation used the wrong acreage weights when combining observations from irrigated and non-irrigated cornfields. The distinction matters because the two types of acreage are not interchangeable when a state yield is calculated. A field sample may accurately describe the plot where it was taken, but an incorrect weight can give one group of observations too much influence and another too little in the resulting estimate.

The agency has not publicly specified why the incorrect weights entered the September calculation. It has identified a processing problem rather than alleging that farmers supplied bad information or that survey workers recorded false measurements. Nor did it assign the error to a particular software system or individual in its notices. Those details should not be inferred from the existence of a correction.

USDA’s objective yield surveys help explain how the error could reach several published statistics. Field enumerators visit sample plots during the growing season, record plant populations and count ears of corn as the plants develop. They return to many plots later, collecting measurements and samples that improve yield estimates as the crop approaches harvest. The collected observations then have to be converted into statistics that represent much larger areas.

In this case, the faulty weighting affected September plant population and ear-count figures for Kansas and Nebraska, as well as the published averages across the ten states covered by those objective yield measures. Yield forecasts for Kansas, Nebraska and the country also changed. The earlier October 23 notice said September corn production estimates were affected at the state and national levels.

The impact therefore extended beyond the two state yield figures. Analysts comparing reported ear counts or plant populations with previous months were also looking at September statistics that required correction. NASS says it has updated the affected Crop Production material and its Quick Stats database, the agency’s public source for agricultural estimates.

Revised September estimates change the comparison with October

The previously published September yields were 122 bushels per acre in Kansas, 177 in Nebraska and 178.5 for the United States. After the correction, Kansas moved lower while Nebraska moved higher. The national average increased slightly because a U.S. estimate reflects production across many states, not an equal-weighted average of state yields.

Nebraska offers the clearest example of why the timing of the revision matters. The October 9 forecast placed its corn yield at 188 bushels per acre. Measured against the old September figure of 177, that appeared to be an eleven-bushel increase. Against the corrected September figure of 182, the difference is six bushels. The correction narrows the apparent month-to-month change without reducing the October estimate itself.

The same distinction applies nationally. The October 9 Crop Production briefing showed a U.S. corn yield forecast of 181.2 bushels per acre and production of approximately 16.03 billion bushels. Using the corrected September yield of 178.8, the October forecast is 2.4 bushels higher, rather than 2.7 bushels above the original September estimate. Those comparisons describe changes in forecasts released at different points in the season; they do not mean yields physically rose by that amount during the month.

For farmers, grain buyers and others using USDA reports, September’s revised historical forecast and October’s unchanged outlook must be interpreted together. Simply keeping the October figure while continuing to use an older copy of the September data would produce a misleading account of how USDA’s expectations changed. The relevant corrected series is now available from the agency.

NASS said it re-examined the October data and estimating procedures and validated the October yield and production forecasts. That does not make the October numbers final harvest results. USDA crop forecasts can still change as more field observations, harvesting information and later survey results become available. The agency’s statement addresses this identified processing problem, not every possible source of forecasting uncertainty.

USDA plans additional checks before publication

The episode has also drawn attention to the controls behind statistics that are widely used in agricultural planning. NASS reports provide common benchmarks for comparing expected supply across regions and months. A change to a previously published estimate can alter calculations used by crop advisers, grain handlers or analysts even when the latest month’s figures are left intact.

The agency says it will introduce additional pre-publication checks to confirm acreage weights and validate objective yield calculations. It is also evaluating other improvements to quality-control procedures. The October 24 notice does not give a completion date or a detailed account of how the new checks will operate, so their effectiveness cannot yet be assessed from the announcement alone.

NASS thanked farmers and other data users for raising concerns that led to the review. That sequence is important: the review followed questions about the October report, yet the specific error USDA identified was in September processing. The correction should not be described as a withdrawal of the October forecast, and the agency did not announce one.

The next scheduled monthly Crop Production report is due November 10 at noon Eastern Time. It will provide another regularly scheduled set of crop estimates against which the corrected September data and unchanged October outlook can be compared.