Baumol Cost Disease Is Killing the Cows
Speculation on an unacknowledged factor in the rising price of beef.
The Counterpoint is a newsletter that uses both analytic and holistic thinking to examine the wider world. My goal is that you find it ‘worth reading’ rather than it necessarily ‘being right.’ Expect semi-regular essays on a variety of topics. I appreciate any and all sharing or subscriptions, as it helps support both my family and farm.
You might’ve noticed the rising cost of beef at the grocery store. The average price of ground beef has risen from $3.87 in February 2020 to $6.54 in November 2025, a ~70% increase in a slightly under six years.
Part of the issue is that the total US cattle herd has been shrinking for decades, with the current herd being the smallest on record since 1952 (below, top). Even if limited to only beef cattle, the same trend holds, though the equivalent-sized-herd year moves up to the early 1960s (below, bottom).
Now, every industry’s dynamics and prices are much more complex than you might first think they are. This is not an attempt to reduce the price of beef to the single explanation. For example, the general bolus of post-pandemic inflation, the demographics of ranchers and cattlemen, increasing amounts of beef imports, and substitution effects of different proteins are all certainly factors in prices of both cattle and beef1.
But one factor that I’ve never seen mentioned with respect to the cattle herd and beef prices is Baumol’s Cost Disease.
Baumol’s Cost Disease
Baumol’s Cost Disease is a theory explaining why wages (and thus prices) rise in labor-intensive, low-productivity fields (e.g. healthcare, education, etc.) as fast as in high-productivity fields (e.g. manufactured goods, technology, etc.). This is because workers in stagnant industries must earn comparable wages to those in productive ones, or else they will switch industries with higher wages, and the resulting labor scarcity will just drive up wages anyway. This drives up prices in low-productivity fields despite little to no productivity gains.
Some specific examples might clarify:
Nurses in the ICU are assigned two patients. This has been the standard for decades. Because of advancements in medical technologies and care, it’s not that there have been zero productivity gains (today’s nurses can care for you better), but that, whether it is 1976 or 2026, if an ICU has twenty patients, then they need ten nurses. There is a fixed ratio of labor to consumer.
Teachers are generally assigned ~15-20 students. Again, it’s not that there have been zero productivity gains (better teaching methods, access to technologies, etc.), but that there is a relatively fixed ratio of labor to consumer.
Musicians might be the quintessential example, because whether it is 1726 or 2026, a Bach concerto takes the same amount of time to play. There has been, quite literally, zero productivity gains. Here, there is an additional constraint: time. It simply takes one hour to play an hour piece.
Compare these fields to, say, a software engineering team within Amazon or Google, who with a single update, could optimize a process that saves tens of millions of dollars, or a manufacturing plant that automates an assembly line in a manner that improves production while reducing labor. The large productivity gains in industries like these lead to higher wages for their workers.
But if wages in fields with low productivity gains don’t keep up with the high productivity industries, then fewer people would enter and/or stay in them. But we need nurses, teachers, and musicians, so that scarcity naturally pushes by wages despite the lack of productivity gains.
Thus we end up with the infamous graph:
Ecology Does Work Too
What is interesting is that I’ve never seen Baumol’s Cost Disease linked to beef cattle, even though both type of Baumol-causing constraints, labor to consumer and time, apply to cattle and beef. Whenever Baumol’s is discussed, the nurses, teachers, and musicians examples are cited and it’s just kind of left at that.
But I think that this is an important discussion to have because people do not generally think of ecology and its ecosystem services as doing labor. Yet that is exactly what they do. Grasslands converting sunlight into carbohydrates and sheep converting grasslands into wool are both just as much labor as a person knitting that wool into a sweater for sale. It’s just a difference branch of the tree of life, plants vs. mammals vs. Homo Sapiens, that are doing that labor.
Cattle Are Low-Productivity
You might’ve seen this famous image of the size of broiler chickens at two months of age in various years. Well, let me assure you that cattle aren’t ~4.5x bigger than they were in 1957.
This isn’t to say that the productivity of beef cattle hasn’t improved. Selective breeding has resulted in cattle that are ~30% heavier than in 1970, better understanding of cattle digestion and nutrition has lead to better feed conversation ratios in feedlots, and improvements in grazing management has lead to better quality and more productive pastures. It’s simply harder to generate marginal productivity gains in larger animals, limiting the increase to the number of consumers that can be fed per cattle head (the constraint of labor to consumer, just like the nurses or teachers per student).
Moreover, compared to R-selected species like chickens and pigs, cattle have relatively long gestations that result in usually only a single offspring, limiting how quickly the size of the herd can change (this is the constraint on time, just like the musicians).
When I asked Gemini2 about the productivity increases of various agriculture products since 1980, beef had the smallest increase.
More Productive Sectors Are Reducing Farmland
Cattle aren’t just facing intra-agriculture competition, but all farmland is facing competition from other more productive sectors. Using the USDA’s Census of Agriculture (taken every five years), I charted the change in acreage of farmland over my lifetime:
~65,000,000 acres of farmland have been lost during my lifetime, ~18,000,000 acres of which were pasture. While that may seem small compared to the ~53,000,000 acres of cropland lost, keep in mind the intra-agriculture productivity gains from the above graph. If adjusted for that, the United States is rapidly losing pasture on a relative basis.
And what are these more productive sectors that are reducing farmland. The biggest four are urban sprawl, conservation lands, data centers, and solar farms.
Urban sprawl is a well-known problem in the United States. Below is an image comparing Atlanta to Barcelona in 1990. Think of the tens, perhaps hundreds, of millions of acres that has been developed across all metro areas of the United States since 1990. For another specific example, let’s focus on Texas, since it’s a well-known cattle state. Consider the growth of the Texas Triangle, not just post-pandemic, but since 1990, and all the acres lost to urbanization and its expansion.
Source: StreetBlog USA ~2% of the entire United States has been placed into conservation over my lifetime. This is through expansion of current National and State Parks, new National and State Parks, private philanthropy projects, public and private environmental works, etc. For example, in 2023, Texas voters approved a $1B spending package for the state to buy more public lands.
Source: Center for American Progress The construction boom in data centers is one of the topics de jour. The number of data centers is expected to double in the coming years, with some being the size of Manhattan. Frederick County, Maryland (where I live) just approved a ~2,600 acre data center zoning area. Assuming ~3,000 current projects at an average size of ~1,000 acres, means 3,000,000 new acres to be consumed in the coming years. This doesn’t include currently operational projects or any of the additional growth that will be needed in coming years.
Source: Axios You’ve probably seen utility-scale solar farms being built in your area. New projects are often built on agricultural land, with pastures being just as affected as cropland.
Source: USDA
This isn’t to say that any one of these industries is “the cause” of farmland or pastureland loss. For example, all solar fields are only ~0.5% of the total area of the United States.
But when you add them up, ~2% to urban sprawl, ~2% to conservation lands, ~0.25% to utility-scale solar projects, ~0.25% to data centers, ~0.25% to miscellaneous projects like new airports and private hunting grounds (K-shaped economy anyone?), suddenly you get significant challenges to cattle grazing and all of agriculture from more productive sectors.
As I’ve said, this essay hasn’t been an attempt to reduce the dynamics of the cattle and beef industries and their prices to this single explanation of Baumol’s Cost Disease, because it is certainly more complex than that. Moreover, I’m just a random philosopher-farmer, not an economist; maybe this isn’t relevant at all.
But when I see a decades-long decline in the size of the American cattle herd, paired with decades-long lags in the productivity of the labor and time intensive cattle ranching, plus competition from multiple “high-productivity” sectors like urban sprawl and data centers, “Baumol’s Cost Disease” was one of the things that popped into my mind.
Yet I couldn’t find anyone that has linked Baumol’s to the cattle industry. As I’ve said, that is probably because people don’t think of ecology and land performing labor, even though our entire existence depends on ecosystems and their services
With no end in sight to this competition from other sectors, I expect beef prices to continue to climb.
To be clear, “cattle” here refer to live animals and their price, which varies depending on the exact animal (e.g. steers (castrated males), heifers (females who haven’t given birth), bred heifers (a pregnant heifer), and cow-calf pairs (a female actively nursing a calf) all have different prices). “Beef” refers to the finished edible consumer product. The price of cattle and the price of beef are independent variables.
Long time readers may have picked up on this vibe; I am a mild-AI “skeptic.” Not in the sense that I think that is an existential risk, nor in the sense that I don’t believe it is capable, but more in the sense of Miyazaki’s famous quote, “I strongly feel that this is an insult to life itself,” that we are opening a Pandora’s box and that we are wholly unprepared for the socioeconomic and cultural harms that will be unleashed.
Believe it or not, I have literally never used AI for The Counterpoint, and only a handful of times in my personal life. But just the other week, I had my first ‘moth flies toward the flame’ moment.
Two of my toddler daughter’s obsessions are unicorns and Pixar’s Cars. So when it came time to make invitations for her third birthday, we tried making invitations with both ChatGPT and Gemini. In short, the results were stunning. Since then I’ve been putting my toes in the proverbial AI-water.
Since this is the first time I’ve used AI for The Counterpoint, I was to establish two promises that I will keep to you. First, I will never use AI for my writing, not even proofreading. I don’t care if it would “improve” my writing; I want you to know that this is written by a human and all his flaws. Second, I will clearly and explicitly label when I do use AI, with exact prompts included as much as possible.
Unless you use natural service for your herd, you have used AI, just saying.
Sheep farmers are exploiting the pasture which grows between and under the panels on solar farms. Unfortunately, not an option for the much larger beef cattle.