Why Michael Pollan is Wrong About Junk Food and Farm Subsidies


. [Please consider supporting Food and Farm Discussion Lab with an  ongoing contribution of $1, $2, $3, $5 or $10 a month on Patreon. Or make a one time donation via PayPal. ]

Marc BrazeauMarc Brazeau | Editor | Food and Farm Discussion Lab | @eatcookwrite

Tamar Haspel
Tamar Haspel (Photo: Gretchen Ertl/Washington Post)

Our friend Tamar Haspel has piece up on The Washington Post about the impact of crop subsidies on making junk food cheap that’s been justifiably getting a lot of attention. She makes a case that most people will find counter-intuitive, if not downright outlandish: Ag subsidies aren’t a significant driver in making junk food cheap.

Michael Pollan is one of the major voices that has been pushing the idea that crop subsidies make junk food cheap, so she gives him a chance to respond. I’m going to gather some of the points I’ve made on Twitter in a lively discussion we’ve been having and some that haven’t made into that conversation yet. This is an area where I think there are huge gaps in understanding and misconceptions that are getting in the way of clear thinking about both US farm policy and public health nutrition interventions.

Before we get started, let me say that I’m not defending our current system of crop subsidies. I’m not a fan for lots of reasons. But pointing out that the most popular criticism is wrong, isn’t a defense, it’s a path to a more productive critique. As Tamar said in her article:

My oyster farm gives me only a limited window into farming; neither I nor Pollan grows carrots or wheat for a living. We have to earn our place at the policy discussion table by getting these things right.

Now, lets take a look at some of the arguments in the piece.

Framing the question:

Pollan, whose work I’m a fan of, held up a package of Twinkies (which cost 99 cents) and a bunch of carrots ($2.99). The Twinkies are a complex food with 39 ingredients, and the carrots are “a very simple bunch of roots,” he said. So why do the carrots cost so much more?

“This perverse state of affairs is not the result of the free market,” said Pollan,… “It has more to do, in fact, with the farm bill.”

The response:

The idea that wholesome foods are expensive and junk foods are cheap because of the system of subsidies in the farm bill pervades the conversation about food policy. But that idea has one very big problem. It’s false.

… Yes, junky food ingredients get much more subsidy money than fruits and vegetables. And I’ve argued against a system that has taxpayers subsidizing foods that are worse for us, rather than those that are more nutritious. But here’s the key overlooked fact: Produce is inherently much more expensive to grow than grains, and that difference dwarfs the difference in subsidy levels.

Crunching some numbers:

A serving of raw broccoli (1 cup, chopped) costs 14 cents to grow. The same size serving of bell peppers costs 9 cents. A cup of strawberries costs 32 cents; blackberries come in at 74 cents. (I was, alas, unable to find current production costs for carrots, although it was not for lack of trying.) Keep in mind that this is the cost to grow and harvest them. They also have to be shipped and stored (and kept cold all the while), and everyone along the food chain gets a cut. By the time they get to you, they’re more expensive.

Now let’s look at commodity crops. The first ingredient in Twinkies is wheat, and a 1-ounce serving of it (enough to make a slice of bread) costs about half a cent. The fourth ingredient in Twinkies is corn syrup, and corn also rolls in at a half-cent per serving.

… Now consider the impact subsidies have. If we take corn, soy and wheat, the three biggest food crops, the 2016 subsidies came to about 10 percent of the value of the crop.

According to supermarket expert Phil Lempert, the actual cost of the food ingredients in processed foods like Twinkies hovers somewhere south of 15 percent, which means that subsidies account for something like a penny — maybe a penny and a half — of those 99-cent Twinkies. And the carrots? A 10 percent subsidy of the value of the crop would be about 3 cents a pound.

So, if we redirect our subsidies away from commodities and toward fruits and vegetables, the Twinkies would cost a penny or two more and the carrots a few pennies less — assuming those subsidies get passed on to consumers and not simply absorbed along the food chain.

Pollan responds:

When I ran that math by Pollan, he warned me: “Don’t lose the forest for the carrots!” Because the bigger issue isn’t the penny or two for the Twinkies, but the long history of subsidies that has encouraged not just the growing but also the improving of commodities. That adds up to huge supplies, which, as any Econ 101 student knows, drive prices down.

And here’s the crucial question: By how much? Pollan and I agree on the mechanism, but disagree on the magnitude.

I think it’s small because I’ve spoken with and read quite a few economists who say so. Daniel Sumner and Julian Alston, economists at the University of California at Davis, have done a lot of work here, and Sumner says subsidies have made corn and soy “slightly cheaper.” Alston estimates that we can attribute 8 to 14 extra calories per day to farm subsidies.

MIchael Pollan: wrong about crop subisidies and junk food.

[Please consider supporting Food and Farm Discussion Lab with an  ongoing contribution of $1, $2, $3, $5 or $10 a month on Patreon. Or make a one time donation via PayPal. ]

CORN: I'm in everything
I don’t remember where I found this gem, but many thanks to whomever made it.

OK, so I think Tamar makes the better case. If you look at the economic literature on the impact of crop subsidies – which we’ve gathered on our research WIKI – the consensus from ag economists is pretty clear – crop subsidies don’t significantly impact cropping decisions or consumer prices.

The biggest chunk of  the farm subsidy budget is for crop insurance and price supports. The research literature finds very little impact on consumer prices. But those aren’t the only ag policies that might affect prices. There are a bunch that tend to increase prices, so as long as we’re slaying the deep seated and persistent myth that the US has a farm policy stance towards cheap food it’s worth running through them.

1. ETHANOL: The Renewable Fuel Standard mandates that millions of acres of corn are used for making ethanol. The impact on food prices is to drive them up as mandated use of corn for fuel increases demand for corn and puts corn for food in competition for corn that the government as mandated must be used for ethanol.

2. SUGAR TARIFFS: Most of American’s protectionist trade policies have been abandoned, but this one remains. Imported sugar is more expensive because of US trade policy. But that also insulates domestic sugarcane and sugar beet producers from competition as well as the manufacturers of high fructose corn syrup (and the corn is more expensive because of the RFS -see #1).

3. CONSERVATION PROGRAMS: The USDA pays farmers to take environmentally sensitive land out of cultivation, especially around water sources and waterways. While enrollment of those programs goes down when commodity prices go up, the net effect is to shrink the supply and drive prices up a bit.

4. LOW INCOME NUTRITION PROGRAMS: Food consumption is one of the few sectors in the economy that benefits explicitly from counter-cyclical fiscal policy. When we go into recession (or say, a financial industry inspired economic crash) enrollment in SNAP (food stamps) and WIC (grocery purchasing support for low income Women, Infants, and Children). These programs first and foremost help enrollees, but a knock-on effect is to nudge food prices a bit higher during recessions than if demand for food was allowed to collapse among low income Americans.

5. MARKETING PROGRAMS: The goal of all the producer Checkoff marketing programs, as well as the USDA Organic certification program is what economists call “demand expansion”, and the goal is higher prices.

That’s just off the top of my head. But the idea that the primary aim of US food policy is cheap food, doesn’t square with the purposes of a lot of programs.

But Michael Pollan’s hypothesis that subsidies have worked in subtler ways over the long haul is also falsified by other fairly straightforward observations about the commodity crop programs. The main being that soybeans weren’t included in the main subsidy programs until 1998, AFTER they had already vanquished oats as the main companion to corn in Corn Belt rotations.

Bear in mind that oats were included in the major commodity programs while unsubsidized soybeans were shoving them to the sidelines. Oats just became a lot less useful after we stopped relying on horses in favor of gas combustion engines.

EWG Soybean subsidies 1995-2016

Iowa Cropping History: 1923-2013
Image courtesy of Steve Savage


FUN FACT: Oats were the original biofuel.

Also notice that the acreage for corn in Iowa has been fairly stable from the 1920s, long before we started subsidizing corn. And if we look at the gains in yield, those started in the 1920s and the trend line changes and then remains stable with each breakthrough in breeding technique. The changes in the rate of improvement don’t correspond to changes in farm policy, they correspond to technological breakthroughs.

Historic Corn Yields in selected regions

Somehow, the “Commodity programs make food cheap and create path dependence.” storyline is only applied to corn and soy. We never hear that complaint applied to barley, rice, oats, flax seed, canola, safflower, dried peas, mustard seeds, or peanuts – all of which are eligible for all or most of the same subsidies that corn and soy are eligible for.

We never hear complaints of our being awash in a sea of cheap subsidized barley and flax. Nor did it take special subsidies for potatoes to dominate the vegetable sector. It just took deep fryers.

If subsidies are warping prices and markets for corn and soy, then why doesn’t that apply to all the other subsidized crops? Maybe because that explanation doesn’t hold water. I think Mr. Pollan has the arrow of causality pointed the wrong way. The bulk of food crop subsidies go to corn because we grow a shit ton of corn, not the other way around.

I also think Michael Pollan is stuck on that story because he has a lot of reputational capital bound up in it. So, the more that story is debunked, the farther he’ll zoom out to increasingly “nuanced”, some might say “eye squinty”, explanations for why it’s true even in the absence of evidence.

So, if it isn’t farm policy making junk food cheap, what is?

This is a story about uneven productivity gains

The reason commodity crops have decreased in price has a lot more to do with fact that commodity crop production has made bigger gains in productivity than specialty crops. Increases in productivity swamp any other effects.

total factory productivity in US agriculture

What that chart shows is that farms are turning out much more product given the same amount of inputs since 1948, with an ugly period from the early 70s to the early 80s, when the charge of “chemical farming” was much more warranted.

While there have been big productivity increases in planting, growing, and harvesting fresh produce, the increase in commodity production has been absolutely massive. That’s because it’s so easily automated and corn is a plant uniquely amenable to improvement through breeding.

Consider one example that crossed my radar in the last week. Matthew Silgar, a California rice farmer did a video on his YouTube channel with old VHS footage of his father’s 1989 rice harvest. In one segment towards the end he compares the productivity of his dad’s combines to this own.

Comparing combine productivity

The three combines he owns today are much more efficient than the three combines his dad owned in 1989.

In 1989, his dad’s three combines could cover 35-40 acres a day, while today, his combines can cover 80-90 acres a day. But the jaw dropping factoid in the story is that a current, state of the art combine could cover 75 acres a day all by itself. For those keeping score at home, that’s 225 acres a day if you had three of them. Or 5½ times more ground in a day than in 1989.

In vegetable harvesting, picking tomatoes for the canned tomato market is highly mechanized. You can see why canned tomatoes are super cheap even though they don’t get the subsidies that corn and soy barley and flax get.

But, even a highly mechanized radish harvest is slow going and labor intensive. And radishes don’t require particularly careful handling. Despite increased mechanization in the field, lettuces still need to be hand picked. There’s no comparison to the way a combine can harvest corn. And we’re only a few years away from a single person being able to operate multiple self-driving combines at once.

At the Intersection of Biology and Chemistry

Quite rightly, Patrick Whittle the food system reporter for the AP in New England demanded a scapegoat if farm policy wasn’t to blame. I suggested the intersection of chemistry and biology.

I replied: You can blame the powerful triple chemical bond of the nitrogen in the air. Biologically, protein is expensive, but carbs are cheap, plants pull carbon out of the air very easily. Nitrogen, not so much.

Photosynthesis (Wikicommons)

This is probably worth breaking down a bit more.

Our air is 75% nitrogen. N2 molecules are super abundant, but they are held together very, very tightly. They are unavailable to the metabolisms of most plants. On the other hand, all plants, and corn more than most plants, pull carbon dioxide from the air and convert the carbon into biomass very easily. This is achieved through an obscure process known as photosynthesis.

Amino acids are the building blocks of proteins. The “amine” in amino acid is NH2. To bring nitrogen into our food, we either need plants that fix their own nitrogen – mostly pulses that yield less per acre than carbohydrate rich crops like corn or potatoes, because the process of nitrogen fixation is biologically complex, or we feed grasses and crops to animals and they synthesize plant proteins and amino acids into protein rich muscles. And a lot of that nitrogen comes from synthetic fertilizer the result of the Haber Bosch process where intense heat and pressure is applied to air and natural gas  synthesize that N2 into ammonium nitrate, a chemically and physically expensive process. 

It’s really easy to turn soil, sun, and rain into a potato or tortilla chip. Making a burger or even a block of tofu takes a lot more resources. So either way, biologically and thus economically, carbs are cheap and protein is expensive.

Getting these things right does matter

We’ve got a lot of problems in our food system. We eat too much junk food and it shows in our health outcomes. The environmental impact of food production should be a lot lower, especially in terms of greenhouse gas emissions, water use, and water quality.

But for people seeking reform and improvement on those metrics, we’ve got limited political capital, limited mental bandwidth, and limited opportunities. We’re never going to skin that cat, if we’re barking up the wrong tree (or mixing the wrong metaphors).

[Please consider supporting Food and Farm Discussion Lab with an  ongoing contribution of $1, $2, $3, $5 or $10 a month on Patreon. Or make a one time donation via PayPal. ]


Please consider supporting FAFDL.org by ongoing contribution of $1, $2, $3, $5 or $10 a month on Patreon.
Become a patron at Patreon!