Though it has rained in large parts of the Corn Belt since August 3, Friday's USDA Crop Production report only includes data from before those rains.

On Friday, the USDA will release its August Crop Production and Supply/Demand report at 7:30 AM CST.

The August production numbers usually cause a stir in the markets because it's the first time the government estimates the crop size with in-field data. In the past, reaction from farmers has been highly critical of the report's accuracy, due mainly to the government's reading of such a young crop.

Continue article

ADVERTISEMENT

The report is influenced by subjective information from farmer surveys conducted following the June Acreage report, and the objective data from in-field samples from corn and soybean crops.

Understanding how the numbers are put together can help temper the criticism that the numbers are inaccurate. Or at least, taking a look at the scientific method the government uses can bring the crop estimates into better focus.

Joe Prusacki, Director USDA/National Agricultural Statistics Service, Iowa Field Office, former Branch Chief for Field Crops in Washington, D.C., told Agriculture Online there is no doubt the estimates become more accurate as harvest draws closer.

"In October, we are getting real fruit, real ears of corn we are harvesting, and very good estimates of what the grain weight is," Prusacki said. "In August, we are still modeling, all we have is a measure of ears. And we're saying, based on history, this many ears should end up with this much weight."

Prusacki added, "What we don't know is what kind of weather we will have from August 1 to harvest. The past few weeks is a good example. The Corn Belt has had good rains, but we couldn't measure that on July 28, when we took these samples."

The same can be said for the soybean crop, Prusacki said.

"Last week it was dry in some areas and the crop was just hanging out there. But, this week, we have been getting rain and cool weather. The soybeans may be saying let's put more laterals and more pods on. That's why guessing this crop is tough in August."

In the past five years, USDA's August estimate underestimated the final crop in all five years. Prusacki said that statistic highlights the difficulty of measuring crop size in August.

"A lot can happen from now until October. For instance, look at 2004, a year marked by very cool summer months and plentiful moisture. A lot of people were worried about an early frost and that the crop wasn't going to make it. The USDA came out with a hefty corn yield and there were a lot of naysayers."

It was cool in July and August. But, September was the sixth warmest on record. As a result, the corn ears put on large amounts of grain, more ears made it to harvest that historically don't, and a big crop ensued.

In 2005, USDA's August count of cornstalks and ears didn't change much to September.

"In addition to what we've already talked about," Prusacki said," Keep in mind that a lot of genetic changes in the corn and soybean seeds have happened since 2000. That can make this crop change as well."

So, what is the scientific method of NASS's crop collection data?

The explanation is in the last few pages of the Crop Production reports, but since few people dig that far, a summarized version includes:

• The August report data is collected from July 28-August 3.

• For the August report, numerous NASS enumerators (a.k.a. crop data collectors) gather 932 soybean samples and 1,045 corn samples in 10 different states that make up 85% of the U.S. crop production.
  "The Sample Design Unit assures us the sample sizes are sufficient to produce statistics that are "statistically defensible," Prusacki said.

• The sample plots are randomly selected in a field. That same spot will be read up to four times before the crop is harvested.

• "The idea is to get a measure of fruit count and based on history what the fruit is going to weigh," Prusacki said. "What we are finding is some years there are more beans in the pods while in others there may be less."

• Prusacki wanted to make clear the fact that crop forecasts are not guesses, influenced by political pressures, or intended to hurt producers.

• "The more producers who get involved with our crop surveys, the more accurate we become. So, we encourage farmers to take part in our collection data by agreeing to fill out crop surveys when we come calling," Prusacki said.

For corn, the enumerators are reading the maturity of the ears in 20-foot of length of row by two rows wide. Also, the data is collected on the kernel length, rows around on the ear, the number of stalks for plant population, the amount of stalks with ears, and the number of ears with formation.

"Historically, there is a relationship between the number of stalks counted in August and the ears found at the end of the year," Prusacki said.

For soybeans, a 3-feet by 3-feet area is measured. The number of plants are counted in that area, and the number of pods per plant.