Sunday, January 08, 2012

What Bt boost?...  

Contra conventional wisdom, new hybrids don't seem to be moving the corn trend line.  In fact, there isn't one right now.

This is less unusual than it might seem.
Statistical analysis finds no statistically significant time trend in average U.S. corn yields since 2003. Thus, the increase in consumption of U.S. corn from 10.2 billion bushels to 12.6 billion bushels between the 2003/04 and 2011/12 crop years, has largely been met through an increase in planted acres of 13.3 million (from 78.6 million acres to 91.9 million acres).
It is important to note that it is not uncommon to find no statistically significant trend in U.S. average yield over a 9 year period. Of all 64 9-year periods starting with the 9 years that begin in 1940, less than half had a statistically significant upward trend in corn yields. This may seem surprising since the average annual increase in U.S. average corn yields since 1940 is 2.2% per year. [More]
It is much harder to come up with some sort of composite (rain, heat, delays, etc.) weather graph that shows a trend is weather influence. But my bet it it would not be flat like the above.

There are many of us with a downward 9-year trend since 2003. Those who have had some spectacular crops in the last few years may be misled into thinking the curve always goes up.

At the very least, the assumption of unlimited productivity advances looks a little shaky to me.

4 comments:

Anonymous said...

Very interesting chart as here(southern ontario)yields have just increased incredibly with province averages surpassing the US average and local farms going from maybe 135 average in 2000 too closer the 200 bu. mark now . I thought it was all the new genetics but by your chart it isn't genetics.regards-kevin

AussieFarmer said...

The explanation may lie in the boost in acreage.

Often when relative profitability changes then field selection sees compromise. So you see more corn on corn, or acreage with poorer corn yield potential is switched from wheat to corn.

This is a common trend when relative profitability results in a substantial acreage expansion in Australia.

Anonymous said...

Noisy data plotted over a short time frame kills trends. (if one were to exclude the first year you flatten things out massively, exclude the final two and it looks like a convincing trend)

For the same reason you can make a case for no change in global temperature if you look over a short enough time span.

John Phipps said...

anon2:

Fair point. Using the same rules, there is no evidence triple-stacks are accelerating the yield curve.

There may be more than one thing going on, as well. Perhaps we are getting a yield slope bump and at the same time seeing the effects of new weather patterns which would have dragged yields down worse without new technology.

What we should see then, is a spectacular national yield when we get an old-time "good year".