Is the Digital Horse Out of the Barn Door?


Written by George Naylor.

I’ve been farming for 48 years, the last 10 organically.  I helped my uncle farm in the early 1970’s when Earl Butz said “plant fencerow-to-fencerow, the world needs your grain.”  Pasture and hay were plowed up, cows sold, and the corn-soybean rotation became the norm.

We shouldn’t be surprised about the digitalization of agriculture.  IBM created many iterations of digital record keeping and analysis over the years following WWII that made the management of giant modern corporations possible.  Those types of business machines now seem quaint.

The evolution of US agriculture has reached a stage where farming operations have become so large that analysis of digital data (with AI?) for decision making is necessary to manage them.  Modern big farmers not only have to figure out optimum fertilizer, seed, and chemical choices, more and more they have to manage a workforce (capitalist agriculture). Since the evolution has bifurcated livestock production into corn and soybean farming, on the one hand, and corporate CAFO/feedlot production on the other, farmers have no choice but to focus on maximizing corn and soybean production—at any cost.

“Precision agriculture” is the means for collecting data.  A modern combine monitors yields as the combine goes through the field creating maps of where the crop is doing well and where it’s not.  Soil testing has been simplified.  Rather than having to send soil to a soil test lab in a little bags from various spots in the field, a probe samples each place that it’s stuck in the ground and in real time data is transmitted to an agronomy office.  Normally an agronomist would compare fertility in areas in the field with the yield data to come up with fertilizer recommendations for different sections of the field (most of the years I’ve farmed, we applied the same rate to the whole field.)  Now all this data is sorted out by computer, so that the fertilizer application machine prescriptively doles out fertilizer for every 4 acre section as it goes through the field.  A computer analysis will even say that different varieties at different populations (seeds per acre) should be planted in different parts of the field.  A modern planter ($300,000-500,000) will do all that as it plants 60 foot swaths over and over again.

Modern farming equipment often includes data collection tools. This farmer is looking at location and system data from a sprayer vehicle that has been uploaded to an app. Photo courtesy of USDA.

When you think of the variability of soil in the field—fertility, soil types, and drainage, and know that big farmers farm many parcels of land in different counties or states, it’s obvious no person could manage such decision making and application.  If agronomic data becomes public or shared with machinery, chemical, fertilizer, or seed companies to be sold to favorite customers or the highest bidder, it will be only a matter of time before the AI farmer rents land out from under a farmer who doesn’t have such high tech capabilities.  Another aspect is the value of this data to land speculators to limit risk of buying a poor piece of farm ground and knowing what the yield potential is so that the highest possible rent can be obtained.  (Robotic tractors, sprayers, combines and drones fit in here somewhere.)  Risk at the production end and speculative end is to be minimized.

This all didn’t need to happen if the policy tools of the New Deal had been preserved after 1953 when bipartisan farm policy was aimed to let the logic of the market move farmers off the land and create an industrialized food and agricultural system.  This understanding is critically important to explain that the consolidation and environmental damage will only get worse without democratic government intervention.  AND because corporate America intends to perpetuate this extractive system no matter what the social and environmental costs around the world.

The evolution of agriculture to the current destructive system can be explained by the declining value of farm products—what corporate purchasers pay for farm commodities.  One way of explaining this phenomenon is the Prebish-Singer Thesis:  “an economic theory that suggests that the prices of primary goods (such as raw materials and agricultural products) tend to decline relative to the prices of manufactured goods over time.”  This economic gravity, as I call it, is as reliable as physical gravity making an apple fall from a tree. (Economic gravity applies to wages also, and that’s why the New Deal initiated the minimum wage—not increased since 2009!)

Falling farm prices (in real dollars) mean some farmers can no longer make a living and get out of farming.  The remaining farmers will have to farm more land or increase their livestock herds and increase yields—become more efficient—to stay ahead of the game. “Get big or get out” was, and still is, the agribusiness mantra.  Because farmers are competing against each other, the least environmentally sound or humane practices become the norm.  Recently one of the few farmers still farrowing and finishing hogs in Iowa just declared bankruptcy with a debt of over $39 million.  There obviously is no such thing as big enough.

Consolidation has long encouraged farmers to “get big or get out” and plant huge swaths of subsidized monocrops, like soybeans.

As an Iowa State University professor told my wife: “It is the nature of agriculture to consolidate.”  That’s probably true for firms in any sector of the capitalist economy.  Well, I say that it’s the nature of parasites to kill their hosts.  Should we let this obvious phenomenon dictate the nature of our economy, our society, and planet’s health? Or intervene to stop the obviously detrimental evolutionary process?

The aim of New Deal price support policy was to create “fair prices for farmers.”  Nevertheless at the same time, this meant that prices paid by corporate consumers were adjusted upward with inflation.  Farm prices paid by corporate consumers were supported by a price floor (using the non-recourse loan—food security reserve system) so that government payments were unnecessary. All government programs that use government payments still allow corporations to buy farm commodities cheap.  A farmer raising livestock on the land with a sound rotation and responsible use of manure would not have to compete with corporate CAFOs/feedlots that order millions of tons of corn and soybean meal to feed millions of livestock with a few clicks of the computer keyboard.

Look at today’s agriculture—Roundup Ready corn and Enlist or Extend soybeans planted contiguously from Canada to the Gulf of Mexico and from the middle of Nebraska to Pennsylvania.  Biodiversity has been extinguished by the exclusion of all other plant species in these fields.   I know it’s easy to say that our efforts are too little too late.  Closing the barn door after the horse is out.  The question is, can we acquiesce to neoliberal inevitable catastrophe by letting economic gravity run its course, or do we devise a vision of a more sustainable agriculture with functioning rural communities and make concrete policy proposals to get us there?