How did you underestimate my ARC payment?

Dear Art,

The ARC payment ended up being significantly more than your estimated payment in September. How could this be?

Michigan Corn Farmer

Dear Corn Farmer,

Almost all of the emails I received asking why the Kansas State University estimated Agriculture Risk Coverage payments had errors were from farmers who received a payment lower than the K-State estimated ARC payment. In fact, you were the only one who asked about our estimate being lower than your final higher Farm Service Agency payment.

The rest of the farmers who contacted us were ones who received an ARC payment that was less than our estimate. In my defense, if one would click on the national map that shows the difference between our payment estimate and the final FSA ARC payment, K-State was either “close” or correct in the majority of the counties.

The maps are located at www.agmanager.info/ag-policy/arc-co-historcial-payment- maps/20172018-arc-co-payment-comparisons.

We posted our final payment estimates in June. At that point NASS had released no final Market Year Average prices. However, this provided little error in our estimates. The K-State estimated MYA prices used for the estimated ARC payments were slightly lower than the final price.

The K-State wheat price estimate was 2 cents below, corn was 3 cents below, grain sorghum was 1 cent below, and soybeans were 8 cents below the NASS published MYA prices used to estimate ARC payments posted in June. So most of the error in estimating ARC payments was due to errors in the yield estimates.

The reason the final approved payment exceeded our estimated payment for your Michigan county was due mostly to errors in the yield estimate. RMA’s county corn yield for your county was 151.1 bushels, NASS’s county yield was 145.6 bushels, and FSA’s approved county yield was 129 bushels, which is what was used to determine ARC payments.

The FSA’s county yield formula is total NASS published county production divided by (total harvested acres minus failed acres). Failed acres are provided by RMA, but the failed acre number is not made public. The soybean county yield is based on planted acres, i.e., total NASS county soybean production divided by NASS planted acres. Corn and wheat use harvested yield to account for silage and wheat graze out, but then must deduct failed acres.

If there is no NASS published data for a county, then FSA can use RMA data. However, FSA can adjust any of those county yields, including NASS county yields based on their expert opinion from the state committee or with other data.

My guess is that FSA lowered the NASS yield due to either a “large” number of failed acres or FSA used the state committee to lower the yield.

This made the payment much larger than our estimate. K-State doesn’t have any access to that information. USDA has three different county yields via three different agencies. If you count RMA’s ARP crop insurance area plan yields, then there are four county yields, and sometimes the numbers are not even close.

There were counties where all three USDA yields were nearly identical. Rawlins County, Kansas, corn is a good example. This county combines non-irrigated and irrigated yields into an average yield for the county. Most of the acres in this county are non-irrigated. In 2017, the combined Rawlins County corn yield published by FSA was 127 bushels, NASS’s yield was 127 bushels, and RMA’s county yield was 128.3 bushels.

There were also counties where the final FSA yield was higher than our estimated yield based on NASS data, or RMA data if no published NASS yield. For example, we missed on the estimated payment for Grant County, Kansas, corn. There was no NASS yield for this county and NASS data was the primary source that FSA used to determine final county yields for ARC payments.

As a result, the K-State estimated yield used RMA county data. Because this was a combined county, we had to average the RMA irrigated and non-irrigated yields into a combined yield. K-State used a weighted average based on the percent of the total acres that were insured as irrigated and insured as non-irrigated. Using a different data set and combining yields for the estimated ARC payment added to the error.

Using this method, K-State estimated the combined irrigated and non-irrigated county yields at 147.5 bushels. There were no NASS yields for this county. The FSA final yield for a combined Grant county corn yield was 208 bushels. As a result, K-State overestimated the Grant County ARC payment by $70.11 and the payment was zero.

Grant County had by far the largest error. The error was also “large” for Hodgeman and Kiowa counties in southwest Kansas, but in the other direction. K-State underestimated the ARC payment for these two counties. Corn farmers in these two counties received larger payments than K-State’s estimate.

K-State compared the estimates with final ARC payments in all U.S. counties. If readers will select the link to the interactive maps and place their curser over their county, a pop-up window will give the additional yield and payment information for that county.

Thanks for a great question!