The 2012 Census of Agriculture provides a wealth of statistics. However, because there were drought conditions in much of South Dakota during 2012, the information in the Census does not reflect normal agriculture activity. That is why it is beneficial to compare information obtained from the Census to other sources of data. Crop and livestock sales are the most common measure of agriculture activity and both were affected by drought conditions in 2012. Lower crop sales were partially offset by higher crop insurance payments.
The National Agricultural Statistics Service (NASS) annually estimates production, the pace of marketings and stocks. The Economic Research Service (ERS) derives an estimate of state-level sales from production and stocks figures in value added tables. Those figures are also refined by the Bureau of Economic Analysis (BEA) in farm income and expenses tables and further allocated at the county level. The currently available allocations from ERS and BEA were released prior to the 2012 Census.
In the 2012 Census, South Dakota farmers reported agricultural sales of $10.2 billion. Of this total, crops contributed $6.1 billion and livestock provided $4.1 billion. The ERS estimated crop sales at $6.4 billion and livestock sales at $3.6 billion in 2012. The BEA estimated crops sales at $6.5 billion and livestock sales at $4.2 billion in 2012. Both also estimated other farm income at $1.4 billion in 2012, which reflected a sharp increase over 2011 levels because of crop insurance indemnity payments. The 2012 Census reveals lower crop sales than earlier expectations would have suggested. Livestock sales totals also differ, but this may be caused by the way the ERS accounts for the part of sales that does not leave the state. The purchased livestock totals for 2012 are similar for the BEA and the Census.
In 2012 NASS estimated that the market value of all field and miscellaneous crops produced was $7.6 billion in South Dakota. The NASS estimate is based on average crop prices and does not reflect actual revenue levels that farmers might have received as a result of market timing and on-farm feed use. Part of the discrepancy in crops data from different sources is timing. The monthly percent of crops sold is reported by NASS at the state-level. These monthly marketings can be combined into 4-month intervals to summarize the relation between production and sales during a calendar year. Consider how corn was measured. The value of corn produced in 2011 was $3.9 billion in South Dakota, and in 2012 it was $3.6 billion. Some of the corn was used as feed. Regardless, sales in calendar year 2012 consisted of about 60 percent of the crop produced in 2011 and 43 percent of the crop produced in 2012 (figure 1). The 2012 Census has corn sales of $3.1 billion.
Figure 1. South Dakota corn marketings
Another major corrective factor provided by the ERS is the adjustment of inventory levels. These are reported for livestock, but must be computed for crops using quarterly NASS reports of state-level inventories of major crops. The change in on-farm stocks of corn, soybeans and wheat shows another impact of the 2012 drought conditions. From December 1, 2011 to December 1, 2012, there was a large drawdown in stocks in South Dakota, especially of corn (figure 2), due to increased feed demand and much lower production. The ERS reported an inventory adjustment for all crops of -$769 million for 2012. Because stocks for feed use and normal delivery flows resumed in 2013, there were fewer bushels for sale during the year. At the state level, the Census-reported amount spent on purchased feed was $1.3 billion in 2012, more than double the $0.6 billion preliminary estimates made by both ERS and BEA.
Figure 2. South Dakota on-farm grain stocks
Some measures are quite close across sources. In the 2012 Census, South Dakota farmers reported receiving government payments in the amount of $283,797,000. This compares closely to the ERS estimate of $330,831,000 for 2012. The ERS totals show a fairly even split between direct payments and conservation program payments. Note that the total excludes any crop insurance payments or premium subsidies.
Both drought and timing concerns show up in insurance statistics. In the 2012 Census, crop and livestock insurance payments totaled $567 million. The large insurance payments were expected given the extreme drought-related crop losses during 2012 in the southeastern part of the state. Separate from the Census, the Risk Management Agency (RMA) reports that the total indemnity payments for crop year 2012 were $1.1 billion, exclusive of livestock. In comparison, the five-year average of indemnities at the state level was $363 million. Thus, the drought impact on insurance payouts was significant.
The disparity in the insurance payments between the 2012 Census and the RMA are substantial. Insurance payments received during calendar year 2012 may reflect crop losses of both the 2011 and 2012 crop years. The RMA periodically reports county level totals during the claim year by crop year. By separating out the likely payments associated with a given crop year, we obtain an estimate for payments received during the calendar year, which is comparable to the Census figures. While this lowers the amount received in 2012 to $938 million, it still exceeds the Census total for the same year. There are two possible explanations for the discrepancy. First, sampling errors and misattributed totals explain some differences. Second, farmers are the only ones receiving indemnity payments reported by the Census -- landowners could have shares insured and received payments and lienholders could also have received payments.
The pattern of the indemnity payments matches the severity of the drought conditions and lower crop values at the county level. Northern counties along the James River Valley received and likely reported payments largely from the 2011 crop year. The largest payments in 2012 were in the southeast part of the state. Hutchinson County had the largest indemnity payments exceeding $100 million for the calendar year. Payment levels exceeded $35 million in Bon Homme, Charles Mix, Clay, Lincoln, McCook, Turner and Yankton counties.
In 2014 the USDA had disaster payments tied to losses in 2012. Payments in South Dakota under the Livestock Forage Disaster Program totaled $66 million. The complex timing of such flows would be difficult to connect back to when the losses occurred from a sector accounting standpoint.
The implication of all of these discrepancies and reconciliations is to use the Census information wisely. The Census provides an excellent view of 2012, but at a point in time that was affected by drought conditions and not necessarily representative of normal production for the state. The impact is particularly pronounced at the individual county level. Unusually low cash receipts in drought-affected counties were partially offset by high insurance indemnity payments. As ERS and BEA revisions occur they will likely show a more accurate picture of the revenue side for 2012, but still not show what would occur under normal conditions.