What is the accuracy of USDA's June Acreage report vs. final acres planted? Iowa State University Extension farm management specialist Steve Johnson looks at that question in the following article. He explains what's happening in 2013 with planted acreage estimates—as USDA is trying to get a better handle on acreage due to the wet spring and delayed planting.
With the record wet spring in Iowa and other areas of the Corn Belt this spring, many acres were planted late and there was a lot of switching from corn to soybeans. Some fields didn't get planted at all--as farmers exercised the "prevented planting" option offered by crop insurance.
During the first 2 weeks of June, data was collected by the
USDA National Agricultural Statistics Service for the annual June Acreage report to be released June 28, 2013. This report combines 2 surveys: June Crops/Stocks, a probability survey that includes a sample of over 71,000 farm operators nationwide including about 2,500 operators in Iowa.
In addition, the June area survey contains about 9,900 segments of land nationwide and over 400 in Iowa. Farm operators from both samples respond to personal and phone interviews along with mail and web-based surveys between May 30 and June 16, 2013. Questions are asked about acres planted and specific crops.
2013 planted acreage number for corn is now forecast at slight increase over last year
According to the survey results released June 28 corn growers nationwide intend to plant 97.4 million acres of corn for all purposes in 2013. That's a slight increase over 2012 and fifth straight year acres have risen. If realized this will be the highest corn planted acreage in the U.S. since 1936 when an estimated 102 million acres were planted.
~~~PAGE_BREAK_HERE~~~NASS also estimates the number of acres of corn to be harvested for grain. Excluded in this number are acres likely to be harvested for seed, food grade, silage, stover and abandoned acres. The balance is then an estimate for harvested acres estimated at 89.1 million acres or 91.5% of acres planted. That's an increase of 1.6% over last year's harvested acres impacted by drought.
NASS survey procedures—a look at how USDA conducts the survey
The June Acreage report validates the Prospective Planting Report released March 28, 2013, but both reports are still estimates of planted acres. The March report is based primarily on farmer planting intentions and is conducted the first two weeks of March. The March survey is a probability survey that includes a sample of over 83,500 farm operators selected from a list of producers that ensures all operations in the U.S. have a chance to be selected.
Statistical methodology and reliability of USDA estimates
The surveys used to make acreage estimates for both the March and June reports are subject to sampling and non-sampling errors common to all surveys. Sampling errors represent the variability between estimates that would result if many different samples were surveyed at the same time. Sampling errors for major row crops such as corn and soybeans are generally between 1% and 3% in March and 1% and 2% for the June report.
Non-sampling errors can't be measured directly. They may occur due to incorrect reporting and/or recording, data omissions or duplications, and errors in processing. To minimize non-sampling errors, NASS uses vigorous quality control in the data collection process and all data are carefully reviewed for consistency and reasonableness.
To evaluate the reliability of acreage estimates in the June Acreage report, the "Root Mean Square Error," a statistical measure based on past performance, is computed. Deviations between the acreage estimates in this report and the final estimates are expressed as a percentage of the final estimates. The average of squared percentage deviations for the latest 20 year period is computed. ~~~PAGE_BREAK_HERE~~~
The square root of the average becomes statistically the "Root Mean Square Error." Probability statements can be made concerning expected differences in the current estimates relative to final end of season estimates, assuming factors affecting this year's estimates aren't different from those influencing recent years. ~~~PAGE_BREAK_HERE~~~
For example, the "Root Mean Square Error" for the corn planted estimate is 0.8%. This means chances are 2 out of 3 that the current corn acreage estimate will not be above or below the final estimate by more than 0.8%
Chances are 9 out of 10 (90% confidence level) that the difference will not exceed 1.4% (margin of error).
Also, shown in the table above is a 20 year record for selected crops showing the difference between the June
Acreage estimates and final estimates. Changes between the intentions estimates and the final estimate released in January during the past 20 years have averaged 532,000 acres. That's a range from 28,000 acres as the smallest to 1.345 million acres as the largest difference.
The Prospective Plantings estimates have been below the final estimate 5 times and above 15 times. This does not imply that the planted estimate for 2013 is likely to understate or overstate the final estimate.
USDA/NASS goes to great lengths to compile accurate information
Conclusion: USDA/NASS works diligently to compile information and provide results of both the March Prospective Planting and June Acreage reports. As a result, the reports are deemed statistically accurate. Note that these numbers are still estimates of USDA's final planted and harvested acreage data to be released January 10, 2014.
However, in 2013 the June survey data was collected before many farmers were done planting due to the wet spring. This is especially true in some northern Corn Belt states such as Iowa, Wisconsin, Minnesota and North Dakota.~~~PAGE_BREAK_HERE~~~
NASS indicated on June 28 that they plan to resurvey farmers in a total of 14 states. This resurvey will be for soybean planted acreage. Some acres, especially corn, farmers in early June likely indicated that they planned to plant. Instead, prevented planting acres were claimed and no crop was planted. Other acres originally intended for corn might have been planted to other crops, primarily soybeans.
USDA to release results of resurvey of soybean planted acres in August Crop Production report
On August 12 USDA/NASS will release results of the resurvey of soybean planted acres in its August Crop Production report. These acreage estimates, both planted and harvested, will be slightly different from the June Acreage numbers. In addition, NASS will release the first in-field yield estimates collected from the June Area Survey data, that's 9,900 segments of land nationwide and over 400 in Iowa.
NASS will also reconcile to the Farm Service Agency the data FSA obtained from farmers at acreage certification. However, the deadline for certifying acres was extended in states including Iowa beyond the original July 15 deadline. So, minor changes in planted acreage estimates for both corn and soybean acres will be reflected beyond the August Crop Production report. The best estimate of corn and soybean planted acreage will not likely be known until the monthly USDA Crop Production reports to be released on September 12 or October 11, 2013.