Last week, the USDA published the year's first data on which crops farmers have planted this year, which allows us to assess the effects of the ongoing drought on crop acres.

The data contain two pieces of information that indicate how the drought is affecting land use. First, farmers report prevented-planting acres, which is land they had intended to plant a crop but were prevented from doing so by a natural disaster, such as inadequate irrigation water. Second, farmers report fallowed or idle acres, which is cropland not planted to a crop.

The reported number of unplanted crop acres in California increased by 130,000 acres from 2020 to 2021. Most of this increase came from prevented planting of cotton and rice. Total unplanted acreage is much lower in 2021 than it was at the height of the previous drought in 2014-15. 

There were almost no 2021 prevented-planting acres in crops other than cotton or rice. Prevented planting reduced cotton acres by 37% and rice acres by 20%. These reductions are similar in magnitude, but larger in percentage terms than the prevented planting in 2015.

These data come from the Farm Service Agency (FSA), the part of USDA that administers farm programs. FSA requires all farmers who participate in crop insurance, farm credit, or disaster programs to report all cropland use on their farm. Thus, they represent only a subset of acres. 

To see how comprehensive the FSA data are, I compared the acreage numbers from FSA to two other sources for several crops using data from our California Crops data app. As I wrote a few weeks ago, the County Ag Commissioners (CDFA) and USDA NASS report similar numbers for some crops and quite different numbers for others. 

All three acreage numbers are quite close for rice, which implies that the FSA's prevented-plant data are a good estimate of how many rice acres the state lost this year. The FSA acreage numbers are less comprehensive for alfalfa and tomatoes, so although FSA shows few prevented plant acres for these crops, it is possible there were acres outside of the FSA umbrella that were taken out of production. 

The FSA planted acres data are even less representative of harvested acres for perennials such as almonds, grapes, and pistachios, so they reveal little about the 2021 responses of  farmers to the drought. 

Another way to measure fallowed or idle acres is USDA's Cropland Data Layer, which is based on satellite images and can be viewed here at our CDL data app. The CDL shows about three times as many fallow or idle acres as the FSA, but similar percentage increases in the previous drought. This may be because the CDL includes fields that are not viable cropland and are not planted in any year, or it incorrectly classifies pasture as fallow. (This is a topic for a future article). We do not yet have 2021 data for the CDL. 

Note: Eagle eyes may note a difference in the CDL data here compared to the graph in my previous article about fallowing fields. I made an error in generating that graph, which I have now corrected.

You may have noticed that FSA records both fallow and idle land.  If you're like me, you wondered what the difference is.  Fallow land is an intentional part of a rotation, e.g., a farmer may plant tomatoes, cotton, and fallow in a three-year cycle, with the fallow year intended to let the field "recover its fertility and conserve moisture for crop production in the next growing season". Idle cropland on the other hand, is just land not planted. It appears that, at least in California, USDA did not report idle land separately before 2017.

So far, the drought has not affected crop acres as much as the 2014-15 drought. Like last time, the reductions are concentrated in the cotton-growing areas of the southern Central Valley and the rice-growing counties of the northern Central Valley.  However, if we get another dry winter, I would expect many more unplanted acres next year. 

But 2021 is not over yet.  Additional water restrictions this year may lead to crop failures, which will show up in the FSA data when it is revised later this year. 

I made the figures using this R code.