by David Widmar
Social media has made communication among producers easier. It has also made it possible for expectations about national corn and soybean yields to be made by simply scrolling. In June and July of this year, it was hard not to conclude the U.S. corn crop was going to be below-trend based on Twitter posts from across the Corn Belt.
While low yields occur in parts of the country each year, how widespread must low-yields be to impact the national yield? This week’s post considers the magnitude of corn acres with low yields in recent years.
To estimate how many acres of corn are affected by low yields, county-level data – yield and acreage – were considered. First, county-level yield data from 1960 to 2016 were de-trended to adjust for the upward trend in corn yields over time. The annual departure from trend-average – or the gap between trend-adjusted yields and the trend average yield – was found for each county from 2000 to 2016. For each year, planted acres of counties with yields of 85% or less of trend-adjusted average were aggregated.
This method has limits. First, it does not account for yield variation within a county. Second, counties were only included in the annual counts if yield and acreage data were both available. Counties missing one of these data points were excluded for the respective year. Even with this limits, the method can provide meaningful insights.
Share of Acres with Low Yields
In Figure 1 is the estimated share of U.S. corn acres with yields below 85% of trends (2000 to 2016). Over this timeframe, 2002 and 2012 standout as the shares were considerably higher. In 2002, 32% of corn acres had county-level yields less than 85% of trends. The share was nearly double – 57% – during the drought of 2012. More broadly, the share has exceeded 10% four times in the past 17 years (2002, 2006, 2011, 2012).
In recent years, the estimated share of acres with low yields has ranged from 2.6% (2014) to 6.2% (2015). In fact, the share has been less than 6.2% in only six years; half of those observations being 2014, 2015, and 2016.
Acres with Low Yields
Looking at the data a little differently, in figure 2 is the number of acres with yields below 85% of trend. Since 2000, the number of corn acres with low-yields has averaged 10.2 million. Excluding the extreme years of 2002 and 2012, low-yields occurred on an average of 6.4 million acres. At 2.2 million acres, 2009 had the fewest low-yielding acres.
At 10.2 million acres, the 17 year average of low-yielding acres is approximately the 2017 corn acreage of Nebraska (9.8 million acres). Excluding 2002 and 2012, the low average of 6.4 million acres is larger than the 2017 corn acreage of Indiana (5.3 million acres). In other words, each year millions of corn acres have low yields. Geographically, the scale of these low-yielding acres can be larger than the acres in an entire state. One must consider this magnitude before making conclusions about the national yield potential for a given year.
What About National Yields?
Shown in Figure 3 is the departure from national trend-yields (2000 to 2016). As expected, years with a high-share of low-yielding corn acres also had a significant, negative departure from national trends (-9.9 bpa in 2002 and -34.6 bpa in 2012). Conversely, years with a small share of low-yielding acres had above trend yields (2004, 2008, 2009, 2014, 2015, 2016).
The correlation between the data from figure 1 and figure 3 is quite strong, -0.94.
Wrapping it up
Each summer speculation about the size of the U.S. corn and soybean crops swings into full-force. Speculation, in both directions, is often supported by anecdotal evidence that is hard to dispute. In recent years, social media has made such evidence more common and more available.
There are many challenges using anecdotal reports when anticipating final U.S. yields and the related market implications.
Millions of corn acres regularly yield below 85% of trend-average. Even eliminating the extremes of 2002 and 2012, the U.S. – on average- has more corn acres with yields below 85% of trend than corn planted in Indiana. Given low yields occur on millions of corn acres annually, this base rate has to be taken into account when thinking about the scale of adverse growing conditions and the possible yield implication.