Behind the Scenes

A look at how crop reports come together

The USDA’s crop production reports draw plenty of attention and plenty of questions about how the data is generated. From March to July, the numbers are created using trend line yields, assumptions about weather, and surveying farmers to find out what they intend to plant and on how many acres. The August report is the first field based survey of the year for corn and soybeans, so the methodology of how the reports are gathered changes.

University of Illinois Agricultural Economist Darrel Good tells Brownfield the USDA uses two survey methods. Starting in late July, around 22,000 farmers in 30 states are asked how many acres of corn and soybeans they expect to harvest and what they expect for final average yields for both crops. The second method has enumerators entering fields in a number of states, about 1,900 fields in 10 states for corn and about 1,800 fields in 11 states for soybeans, counting ears or pods, along with other factors that indicate production potential. Those factors, along with historical data, are then used to forecast average yield numbers.

Good says the two surveys rarely agree, “The farmer survey shows a lower yield than what actually happens and the objective yield or field yield survey typically would show a yield higher than what actually materializes, but when they blend the two, then they feel like they get a pretty good forecast of actual yield. In August, in some years, there’s not much to measure, not much to look at, in other years, there’s a lot. By the nature of the crop maturing and eventually getting to harvest, the accuracy of those numbers certainly improves as you go from August through November.”

The enumerators are hired by the USDA, from the area and have an agricultural background. The farmers surveyed directly are randomly selected respondents from a list of producers who in June indicated they planned to plant corn and/or soybeans that year and then surveyed from August through November. The enumerators are frequently used over the course of several years and new farmers are selected every year.

As far as the reliability of the August crop report, Good says “The forecast in August is not biased. It’s not routinely too high or too low, compared to the final estimate. On average, it’s about equal, but any individual year, there can be and have been large deviations from the August number to the final number. For corn for example, in 2012, there was less than a one bushel change from the August forecast to the final number. In other years on corn, that number can change by as much as 15 bushels per acre. While it’s not biased, it can deviate from the final estimate in any given year.”

 

Add Comment

Your email address will not be published.


 

Stay Up to Date

Subscribe for our newsletter today and receive relevant news straight to your inbox!