The following commentary does not necessarily reflect the views of AgWeb or Farm Journal Media. The opinions expressed below are the author's own.
See the latest reader comments and hear John explain some of agriculture’s complex topics.
Sue Starr sent a note regarding the wheat crop reports:
"How can you accurately predict the 2017 wheat harvest without considering: small number acres of wheat planted; thousands of wheat acres already baled for livestock feed; abandoned acres due to the drought and extremely small yield due to the drought. You are not reporting the whole picture."
Sue, thank you for your feedback. All of us here, and I'm sure our market experts, have heard this criticism not only this year but every year. Some analysts even think USDA attempts to estimate the crop are a mistake entirely.
But what took me a long time to realize is that these crop size numbers are not intended to be the definitive word on how much grain is out there. They are numbers that we get when we do the same yield estimating steps every year.
In-season crop estimates should be weighed against previous estimates, not against absolute accuracy. By looking at the long history of these estimates, farmers and traders can place their own value on their worth. The key is consistency, more than accuracy.
However, as for accuracy, you can find graphs like this to help you decide how well estimators do. This is for the May WASDE wheat yield estimate.
Values above the line mean the final yield was higher than the May estimate. The average was 2 percent higher.
While we all can and do argue with how these estimates are calculated, it really doesn't matter how they do it. What matters is if are they accurate at all.
My thinking is if they do it the same way each time, history can tell us something about that accuracy.
No comments have been posted to this Blog Post