In 5 of the last 10 years, payouts on crop insurance to US cotton farmers exceeded 17% of the market value of production. Most of those payouts were for losses due to drought, but excess rainfall was also a factor.

Indemnities were lower before 2010. Do the high indemnities since 2010 reflect high risk or something else?

I don't have a background in insurance, so I sometimes get lost in the terminology. I find analogies to personal insurance helpful. If we observe high payouts on homeowner insurance as a percent of home value, it could be because a lot of people are insuring their houses (high liability), bad luck (high loss ratio), or because house fires are highly likely (high risk). 

Cotton crop insurance liability has increased. From 1995 to 2010, crop insurance liability on cotton averaged about $3b per year. The 2008 Farm Bill restructured crop insurance, leading to a jump in insurance liability, mostly in the form of revenue protection (RP).  Liability averaged about $5b in the most recent decade.

The increase in insured liability did not come from an increasing value of production. Insuring a certain percentage of a higher production value would entail higher insurance liability, but the value of US cotton production has been relatively constant at around $6b per year for the last two decades.

Farmers taking on more insurance liability explains the trends over time, but it does not explain why insurance payouts seem so high relative to the value of production.

Loss ratios have been consistently higher for cotton than other commodities, averaging 1.15 since 1989. This means that insurers paid out 15% more than they received in premiums on average. For comparison, the loss ratios for corn, soybeans, and wheat were 0.78, 0.65, and 1.04 over the same period.  In the last decade, cotton loss ratios averaged 1.02, which exceeds the values for corn (0.92), soybeans (0.53), and wheat (0.76).

So, insurers have lost more money on cotton than on other major crops, which suggests a 30 year string of bad luck. It would be easy to suggest that premiums need to be higher, but cotton insurance premiums are already high. 

Cotton insurance premiums in the last decade have averaged about 20c for every dollar of liability.  Specifically, premiums have average 15% of the value of production and those premiums have bought liability equal to about 75% of production value, and 15/75 = 20%. Over the same period, premiums were 9c per dollar of liability (7/75) for corn, 10c per dollar of liability (6/60) for soybeans, and 15c (11/75) for wheat.

High insurance premiums tell us that cotton is a relatively risky crop. To understand why, we first note that half of cotton acres are in Texas.  California was the second largest producer in 1980, but acreage in the state has declined 88% in the past 40 years. Georgia is now the second leading state. There is almost no cotton grown in the grain-growing regions in the midwest and great plains.

The high risk to growing cotton reflects conditions in Texas, which has experienced multiple years of drought in the last decade, especially in 2011 when almost 100% of the state was affected. In 2020, the Texas cotton crop was severely affected by storms, including Hurricane Hanna. Last week, the USDA reduced its estimate of the 2020 cotton crop by 1.1 million bales (about 7%).  This pattern suggests that the environment in which farmers grow cotton is a large driver of the risk.

American cotton farmers pay high crop insurance premiums and receive high indemnities, which reflects the high risk of growing cotton relative to other crops. Beyond cotton, it reveals the risk of growing crops in areas prone to drought and other extreme weather, which provides a warning for the future of agriculture in a changing climate.

These plots come from our newly improved Crop Insurance data app and our US Crops data app. Huge kudos to UC Davis ARE PhD student Seunghyun Lee for putting these apps together.

For more on how US crop insurance works, see this Ag Data News article.