Tracking USDA’s Crop-Projection Process

After farmers’ intentions to plant record acreage to soybeans came to light last spring, the question industry analysts tried to answer for most of the growing season wasn’t whether the crop would be big.

It was: How big?

Now at the end of the year, farmers can judge their crop using real numbers for yield, total bushels produced, how many bushels they put in their bins and the price they received for the bushels they delivered to the elevator.

But those numbers aren’t clear in August, when the U.S. Department of Agriculture (USDA) releases its first of four monthly Crop Production reports that attempt to estimate how big the soybean crop will be.

In the November report, the last one issued before Beyond the Bean press time, USDA expected U.S. farmers to harvest 3.96 billion bushels of soybeans, while some private analysts called for the first 4-billion-bushel crop in history. The previous record for U.S. soybean production came in 2009, when farmers harvested 3.36 billion bushels.

These reports have been criticized for their negative effect on prices during abundant years like this one, but Doane Advisory Services Senior Economist Martyn Foreman says they also have advantages.

“Having these hopefully fairly accurate and unbiased estimates helps the markets operate more efficiently,” he explains. “If we think we know the crop size or how many acres will be planted, then the markets will benchmark and adjust accordingly, and it should create less volatility.”

USDA calculates its estimates using a combination of both farmer surveys and in-field measurements taken by trained staff. These methods have remained the same over time, and agency statisticians have learned to make minor adjustments to compensate for slight biases in the data that they know exist.

USDA conducts monthly surveys of a random sample of soybean farmers from 29 states, asking them what they expect their yield to be. It combines those figures with data gathered by a team of technicians who take yield-predicting measurements from random fields throughout the country.

“The primary source of all this data is the producers on the frontline,” says Hubert Hamer, Statistics Division Director of USDA’s National Agricultural Statistics Service (NASS). “We rely on their cooperation to provide the most vital and meaningful part of the estimating program. We also protect their individual identity, as required by law.”

Next, USDA checks and rechecks the data in various ways before a board of seven to 10 statisticians reviews it on a regional basis and determines national yield and production forecasts. The numbers remain literally locked down until their release to everyone at once so that no one has an unfair trading advantage.

“NASS realizes the importance of providing timely and reliable data, remaining free from political influence, and refraining from policy advocacy,” Hamer adds.

Foreman says these projections are scientifically derived, unbiased and have not changed much over the years.

“If farmers understood how extensive the work is and the quality of the people doing it, I think they’d feel a lot more comfortable that it’s done pretty scientifically,” he says.