Guest Post by Madge Waggy
Just after the morning deliveries and before the first customers, a supermarket looks both ordinary and slightly deceptive. The lights are on, the coolers hum, and the shelves hold exactly what they held last week. Strawberries show up in February, lettuce comes from ground that was desert two generations ago, and on the same shelf beef raised across three states sits beside bread baked from grain cut a thousand miles away. Someone pushing a cart through the whole store can finish the errand without once wondering how any of it got there.
That is part of why the American food supply is so easy to take for granted. Nobody praises the trucking fleets, the cold storage, the seed suppliers, or the packing houses, because a thing that rarely fails stops being noticed. Two or three generations have grown up treating food the way they treat running water: as a fact of the world, unremarkable and about as interesting.
Reliability, though, is not the same as resilience, and the two are easy to confuse because the difference only appears when something goes wrong. A farm can still bring in a decent harvest during a dry year, until that dry year coincides with expensive fertilizer, a thin labor crew, and a delayed shipment. Shelves can stay full while carrying fewer brands at higher prices. The country is not running out of food. What it may be running out of is the cushion that once let a bad season pass without anyone noticing.
This article rests on three kinds of claims, and it is worth keeping them apart. First come the demonstrated problems: measured facts about the age of farmers, the drawdown of aquifers, the concentration of meatpacking, the cost of inputs, the price of food. Then comes an interpretation, which is mine and can be argued with: the suggestion that these separate trends, taken together, are eroding the system’s ability to absorb a shock. Last come scenarios, which are conditional rather than predictive, describing what could happen over the next decade or two if several pressures arrive at once and little changes. The text will say which is which as it goes.
What follows is not a forecast of collapse. It is an attempt to look past the grocery aisle at the people who grow the food, the water and soil they depend on, the infrastructure that moves it, and the habits of a public that has rarely had a reason to worry.
American food logistics are organized around movement rather than storage. Warehouses are designed for turnover, distribution centers run on delivery windows that assume trucks arrive when promised, and most supermarkets carry only a few days of stock, though the exact figure varies by product; fresh produce and bakery items turn over within days, while canned and frozen goods sit much longer. Very little slack exists by design, because idle inventory costs money and motion does not.
This is what the industry calls just-in-time, and for decades it worked remarkably well. Waste fell, prices stayed low, and a range of food reached the shelf that no earlier era could match. The bill for that efficiency comes due only under stress, and when it does, the buffers get spent quickly. A single failed truck route can empty a produce section for days, and one plant going dark for repairs can thin meat counters across several states. Fewer alternate paths run through the network than most people assume, and the paths that do exist tend to funnel through the same handful of junctions.
Older food economies leaned on redundancy instead. Grain sat in silos, towns fed themselves, and a surplus worked as insurance against a bad year. In the modern version, surplus reads as waste and storage reads as cost, so redundancy survives mainly where it happens to escape the spreadsheet. What remains runs beautifully when conditions hold and begins to creak the moment several things break at once. The efficiency is real, and so is the fragility that comes with it.
Concentration belongs to the same design. A small number of plants handles an enormous share of the nation’s meat, grain, and packaged food. Four companies — Cargill, JBS, Tyson, and National Beef — slaughter roughly 80 to 85 percent of the country’s fed cattle, a figure the Government Accountability Office put at about 85 percent in a 2021 report and which has moved little since. The top four pork firms account for roughly two-thirds of hog slaughter, and poultry is more concentrated still. On the grain side, four international traders, often nicknamed the ABCD firms, have long handled a large share of world flows, with American exports passing through a shrinking number of elevators and ports.
When the packing sector sneezes, the meat case in every region catches cold. The clearest recent example came in the spring of 2020, when COVID-19 tore through crowded slaughterhouses. A Smithfield pork plant in Sioux Falls, South Dakota, processing something like 4 to 5 percent of the nation’s pork at the time, shut down after becoming one of the largest single-source outbreaks in the country. A Tyson beef plant in Columbus Junction, Iowa, closed. JBS and National Beef idled facilities in Colorado and Kansas. At the peak, beef and pork production fell by roughly a quarter, and the shock moved in both directions at once: prices paid to farmers collapsed, sometimes to the point that hogs were euthanized for lack of processing capacity, while retail meat prices climbed. The same disruption reached producers and consumers from opposite sides, which is what concentration does. It transmits a local failure across the whole chain.
None of this makes consolidation irrational. Real reasons stood behind it: economies of scale, thin margins, decades of mergers, and a regulatory climate that looked kindly on size. A large plant can slaughter more animals per hour at lower cost than a small one, and consumers have benefited through cheaper meat. What the system traded away was its detours, and with them the ability to route around a failure.
Leverage hides inside that structure without anyone seizing it deliberately. It emerges on its own from squeezing production and transport into fewer hands in the name of cost. Abundance becomes conditional rather than guaranteed, and a small number of failures can move food away from millions of tables at once.
Interpretation begins here. It is tempting to say that just-in-time always made the system fragile, but that oversimplifies. Grocery stores have always carried limited shelf stock; a butcher in the 1950s did not keep weeks of beef either. What has changed is the network behind the shelf, with fewer plants, fewer carriers, fewer alternate routes, and thinner reserves at every layer. The modern system is not more fragile because it moves goods. It is more fragile because the number of places where a single failure can stop the flow has grown while the number of ways around those places has shrunk.
Rebuilding redundancy is slow. Storage, local processing, and alternative routes take years to restore once they have been allowed to disappear, and the disappearance happens on a schedule of cost savings rather than of visible shortages, so the consequences arrive long after the decisions that caused them. A regional plant closed in 2015 to save money does not show its cost until a disruption in 2025, by which point the plant is a parking lot.
One wrinkle rarely enters the discussion. Efficiency and resilience are not opposites so much as competing goods, and the trade-off stays invisible because its cost is deferred. Every dollar saved by running leaner shows up immediately on a balance sheet, while every unit of resilience given up materializes only in a bad year. In good years the lean system looks strictly superior, which is why it keeps winning and why the arguments for holding reserve keep losing. The logic is not wrong. It is incomplete, because it prices the present and discounts the future.
That asymmetry explains a great deal about how the system got where it is. Managers are rewarded for the savings they can demonstrate and only occasionally punished for the resilience they quietly surrender. A warehouse that keeps a little extra stock looks wasteful until the day a truck is late, and by then the person who approved the extra stock may have moved on. Over decades these small decisions compound in one direction, and the system ends up leaner than anyone consciously chose.
The same logic runs through transportation. Trucking moves the large majority of American freight by weight, commonly estimated at around 70 percent, and the industry has cycled through driver shortages for years, with trade associations citing gaps in the tens of thousands. Rail, which hauls grain, fertilizer, and other bulk goods, has consolidated down to a handful of Class I carriers, and service complaints spiked in 2022 as the railroads cut crews and lengthened trains in pursuit of their own efficiency. Ports handle imports of everything from coffee to out-of-season produce, and when they clog, as Los Angeles and Long Beach did through 2021, the delay propagates through the whole country.
None of these links is broken. Each is simply running with less give than it once had, and give is exactly what a shock consumes.
The hands working the land are older than they used to be, a pattern any drive through farm country will show and the numbers confirm. According to the 2022 Census of Agriculture, published by the USDA’s National Agricultural Statistics Service, the average age of an American farm producer has reached 58.1 years, and the share past 65 keeps rising. Fewer young people are stepping in behind them.
The reasons are not mysterious. Farming means heavy debt, expensive equipment, volatile input prices, and weather that no longer keeps its word, and a single ruined season can erase years of work. Against a steady city paycheck, that is a hard case to make to someone choosing a career. So when growers retire, the land often fails to pass to new hands. It gets bought by larger operations, cut into subdivisions, leased for something other than food, or simply left idle because keeping it in production no longer pays.
The scale of the shift is easy to understate. The 2022 census counted about 1.9 million farms, down from roughly 6.8 million in 1935, and the long decline has never really reversed. Land in farms, a category that includes pasture and range as well as cropland, totaled about 880 million acres in 2022, down slightly from 2017. Cropland proper is smaller, on the order of 300 million acres, and it matters not to blur the two, since a ranch in Nevada and a cornfield in Iowa are counted under the same headline even though they function very differently. What matters for the food supply is the cropland, and the trend there is gentler than the farm-count trend, though still downward at the edges.
A second pattern sits underneath the averages. Most farms are small, and most production comes from a few of them. Roughly 9 percent of farms, those with sales above one million dollars, account for the bulk of the total value of what American agriculture produces, while the bottom half of farms by sales contribute only a sliver. The median farm household often earns more from off-farm work than from farming itself. This is less a sign of failure than a structural fact: the business of growing food has concentrated into fewer, larger, more capitalized operations, while the small farm survives as a part-time pursuit.
The consequence for resilience is not obvious. A large operation can invest in irrigation, storage, and equipment that a small one cannot afford, which arguably makes it more robust. But the same operation is also more exposed to a single shock, whether a bad commodity price, a failed crop, or an interest-rate move on a large loan, and the loss of many small, diversified farms removes a kind of distributed redundancy that the system once had without anyone counting it.
A second loss draws even less attention, and it is buried inside that shift: knowledge leaves with the people who hold it. When to plant a difficult crop, how to read a particular patch of soil, which pests arrive in which week, how to carry a marginal field through a dry August. None of that lives in a manual. It passes down through generations and goes quiet when a farm changes hands. Machinery can replace some of it and not the rest, and the part it cannot replace is often the part that matters most when conditions turn strange.
The coming handover is substantial. Farmland is old, and a great deal of it will change hands in the next couple of decades as the current generation retires. Estimates vary, but the American Farmland Trust has projected that on the order of several hundred million acres could transition ownership in the next twenty years, and whether that land passes to working farmers or to developers and investors will shape what the countryside looks like. When a farm is sold to a larger neighbor, the land stays in production but the family knowledge often does not transfer with it; when it is sold for houses, the production stops altogether.
Small fruit growers in Washington state offer one illustration. Orchards there have long depended on skilled hand-picking, and older growers perfected a judgment about ripeness, timing, and handling that no machine has fully matched. As those growers retire, the expertise thins, and yields can suffer even though the land itself is unchanged.
Labor enters the picture next. Much of American agriculture runs on seasonal work, and seasonal work answers to biology rather than to anyone’s schedule. Fruit and vegetables ripen inside narrow windows, and missing the window loses the crop no matter how good the growing season was. Labor gaps have already left produce rotting in fields for want of pickers. The work leans heavily on immigrant crews, and a large share of hired farmworkers are foreign-born, with a substantial minority lacking legal status. The estimates vary, but the foreign-born share is well above half in many crops. Any kink in that pipeline turns directly into food left to spoil in the sun.
The H-2A visa program, which brings in temporary agricultural workers, has grown dramatically as a result, from roughly eighty thousand certified positions a little over a decade ago to several hundred thousand today. That growth is itself a signal: American agriculture increasingly cannot staff its harvests from the domestic labor pool alone, and the program that fills the gap is subject to political weather.
Timing is what makes the labor question brittle. A factory can pause a line and restart it later without losing much. A field cannot. Crops do not wait for a policy debate to settle and do not care what the calendar says. A shortage does not simply reduce the harvest; it reduces the harvest at the exact moment the crop is worth the most, because fruit that was perfect on Monday is mush by Friday. Processing plants that expected a delivery sit idle, freight schedules slip, and availability drops just as demand peaks.
Tomatoes must be picked and moved within a short span before they soften. Berries tolerate almost no delay and almost no rough handling. Lettuce and leafy greens are cut in tight windows and spoil quickly. Each depends on crews arriving exactly when they are needed, so any disruption, whether a change in immigration policy, an economic downturn, or a public health emergency, lands directly on harvest timing.
On top of labor sits cost. Modern growing depends on a web of inputs most people never picture: diesel for machines and freight, fertilizer produced through energy-hungry chemistry, spare parts drawn through global supply chains, irrigation tied to water that keeps moving, and packaging hostage to industrial cycles. Fertilizer prices in particular spiked in 2021 and 2022, in some cases doubling or tripling, driven by natural gas costs and the war in Ukraine, since nitrogen fertilizer is essentially made from natural gas. Diesel followed oil, and farm debt has climbed past half a trillion dollars.
When those costs surge, farmers do not usually fold right away. They get careful. They plant fewer acres, walk away from marginal ground, choose crops that demand less labor and carry less risk, and cut output to protect the balance sheet. Every decision is sensible from where the grower stands, and every decision narrows the national supply a little further.
For a customer, the shrinkage stays invisible for a season or two. It accumulates quietly, spread across regions and years, until the gap reaches the shelf and nobody can say where it came from. Four adjustments recur: fewer acres planted, marginal land abandoned, a shift toward low-risk crops with steady demand, and lower overall output to stay solvent. Each move is rational on its own, and collectively they shave both production and resilience at the same time. No farmer sets out to weaken the food supply. The weakening falls out of thousands of private decisions made for private reasons, none of them coordinated, all of them pointing the same way when costs keep rising.
Consumers rarely see one piece of this, and it deserves its own moment because it reframes who pays. The share of the retail food dollar that reaches the farmer is small and has been small for a long time, on the order of fifteen cents by the USDA’s own accounting, with the rest absorbed by processing, packaging, transport, marketing, and retail. When retail prices rise, the increase does not flow evenly back to the farm. During the 2020 meat disruption, the gap between what farmers received and what shoppers paid widened sharply, and much of the spread stayed with the middle of the chain even after production recovered. The point is not that anyone is cheating. It is that the farm is a small and vulnerable node within a long chain, and shocks that look like a food problem from the aisle often register at the farm as a price collapse.
Some of the pressure comes from the ground and the sky, and the mechanisms are worth separating, because they are often lumped together into a vague sense of climate without anyone distinguishing what is actually happening.
Start with water. The Ogallala Aquifer, which underlies parts of eight plains states from South Dakota to Texas, supplies a significant share of American irrigation, often cited at around 30 percent of the nation’s irrigation groundwater, and it has been declining for decades. The decline is uneven: some areas have dropped by more than a hundred feet since large-scale pumping began, while others have been steadier. Kansas, which depends heavily on the aquifer, has recorded some of the steepest losses, and wells have run dry in parts of the Texas Panhandle. The depletion is patient rather than dramatic. Where water becomes uncertain, planting becomes a bet, and land without reliable water cannot reliably feed anyone. Over-extraction leaves its own damage: sinking ground, rising salinity, falling yields.
California’s Central Valley tells a related story with different mechanics. The valley is the core of a state that grows a large share of the nation’s vegetables and tree fruit, and it leans on both surface water and groundwater. During droughts, farmers pump harder, and the aquifer beneath the valley has dropped in some places dramatically, causing the ground itself to sink, with subsidence that has damaged canals and roads. The state’s Sustainable Groundwater Management Act, passed in 2014, requires the most stressed basins to reach sustainability by 2040, which sets up a genuine reckoning: bringing pumping into balance will likely mean taking a substantial amount of land out of irrigated production. That 2040 deadline is one of the few hard dates in this story, and it is worth remembering when the later scenarios arrive.
The Colorado River adds a third case. The river’s flow is allocated among seven states and Mexico under a compact written in 1922, when the river was running higher than it does today, and the allocations exceed the actual supply in most years. A long drought, punctuated by negotiations and emergency cuts, has forced the basin’s users to reduce withdrawals, and the lower basin states have committed to significant temporary reductions. The river irrigates a great deal of winter vegetable production in Arizona and California, so what happens on the Colorado reaches dinner plates across the country.
Soil comes next, and the common shorthand that chemical fertilizer erodes fertility needs a little care. The heavier problem is more physical than chemical. Decades of intensive tillage have stripped organic matter and broken up soil structure, and degraded soil holds less water, resists erosion less well, and recovers more slowly. Fertilizer can mask some of that loss by supplying nutrients directly, which is why yields have stayed high even as the soil underneath has thinned. It is a patch rather than a repair, and ground that has been mined rather than farmed does not recover on its own. Rebuilding it takes cover crops, rotation, and patience, the kind of patient work a system tuned to quarterly results does not always reward.
The numbers here are hard to pin precisely, but the direction is well established. Soil erosion on American cropland runs to roughly one and a half to two billion tons a year by common estimates, most of it tolerable in any given year but cumulative over decades. Cover crops, one of the main remedies, are used on a small minority of cropland, on the order of five percent, though growing. No-till and reduced tillage are more widespread but uneven. The farmer and writer Wendell Berry called soil “the great connector of lives, the source and destination of all,” and the line reads less like poetry and more like a specification once you understand what healthy ground actually does.
The weather itself comes last. Farmers once leaned on patterns inherited across generations, reading rainfall, planting windows, and the rough shape of a normal season from what their parents and grandparents had seen. That inheritance is coming apart, though coming apart is itself a metaphor, and the concrete changes are more useful. From 2000 into the early 2020s, southwestern North America experienced a megadrought that researchers, in a widely cited 2022 study in Nature Climate Change, judged the worst in at least twelve hundred years. Global average temperatures set records in 2023 and again in 2024. The USDA redrew its plant hardiness zone map in 2023, and much of the country shifted warmer. Flash droughts, sudden and intense, have become more common, and the extremes have grown more extreme, with heavy downpours in some places, prolonged heat in others, and the same region sometimes seeing both within a year.
Gradual change is manageable. Volatility is harder, because when the pattern stops being trustworthy, planning turns into guesswork. A grower can adapt to a warmer average by planting earlier or switching varieties. A grower cannot easily adapt to a season that swings from flood to drought and back, because the equipment, the contracts, and the calendar are all built around a rhythm that no longer holds. Heat that once came in predictable waves now lands at odd hours, stressing animals and trimming yields in ways growers struggle to anticipate.
The 2021 Texas freeze, Winter Storm Uri, offered a preview of how these pressures interact. The storm knocked out power and water across the state, froze citrus and other crops, and disrupted food processing and transport, all at once. It was a weather event, but its damage ran through the same channels that every other pressure uses: energy, water, labor, and logistics.
Each factor alone would be survivable. Together they compound, and a system never designed for persistent unpredictability feels the strain in every decision a grower makes. It is also worth noting that climate is not a single lever. A warmer average, a drier aquifer, and a more volatile season are three different problems with three different remedies, and conflating them makes the response vaguer than it needs to be.
Food in this country routinely crosses thousands of miles between field and fork, which requires cheap fuel, available drivers, working highways and rail, and ports that keep flowing. A snag anywhere delays everything downstream. Often the food exists and simply cannot arrive in time, a distinction that matters more than it sounds, because shortages frequently erupt not because nothing was grown but because what was grown could not move. A household staring at a bare shelf has no way to tell the difference.
The idea of food miles is often overstated, since a tomato trucked from Mexico can carry a smaller carbon footprint than one grown in a heated local greenhouse, but distance matters for a different reason: it multiplies the number of things that can go wrong. Every additional mile is another chance for a breakdown, another border, another handoff. A short local chain has fewer failure points by construction, even if each one is not inherently more reliable.
Four weak points carry most of the risk. Fuel spikes make every mile costlier and push marginal routes toward unprofitability. Driver shortages limit how much can be delivered even when the trucks exist, and the trucking industry has cycled through shortages for years, with the American Trucking Associations estimating gaps in the tens of thousands and warning they could grow. Port congestion stalls the imports and exports the domestic system depends on. And aging infrastructure, from bridges to rail lines to terminals older than most of the people using them, raises the odds of a breakdown along a route with no realistic alternate.
The 2021 port crisis is instructive. Backed-up container ships off Los Angeles and Long Beach, driven by a surge in imports and pandemic-era bottlenecks, delayed everything from furniture to coffee and forced retailers to reroute and ration. The ports eventually cleared, but the episode showed how a single chokepoint on the coast can ripple into shelves two thousand miles inland. Rail had its own version of the same problem in 2022, when service deteriorated badly enough that regulators held hearings and shippers complained of missed deliveries and unpredictable schedules.
Cold chain adds a layer most people never consider. Produce, meat, dairy, and frozen goods all depend on an unbroken sequence of refrigerated storage and transport, and a break anywhere, whether a power outage at a warehouse or a failed compressor on a truck, can spoil a whole load. The cold chain is one of the genuine engineering achievements of the modern food system and one of the most brittle, because it has no tolerance for interruption.
The infrastructure itself is old. A large share of American bridges are past their design life, and the rail network, while privately owned and generally well maintained on the main lines, has seen deferred maintenance on the margins. The country has underinvested in the physical layer of its food system for decades, not because anyone decided to, but because maintenance is invisible and unglamorous while new projects are visible and political.
Fuel deserves a special note, because it sits underneath everything else. Diesel powers the trucks, the trains, the tractors, and the ships that move food, and it also powers the fertilizer and the irrigation pumps. When diesel prices spike, as they did in 2022, the cost propagates through every link at once, and it hits the most marginal operators hardest, meaning the small trucking firms, the family farms, and the regional distributors, which is to say exactly the nodes that provide redundancy to the system. A fuel spike does not just make food cost more. It quietly removes some of the smaller players who would otherwise have absorbed a shock.
None of these needs to fail completely to matter. Each only has to slow down, and the delays stack. A driver shortage, a port backup, and a fuel spike in the same quarter do not add up to three problems. They multiply into one much larger one, because the system has no slack to absorb them in sequence.
The softest vulnerability may live inside people’s heads, but it is worth being precise about what the public actually experiences, because the easy claim that Americans have never known scarcity is only partly true and can mislead.
It is true that the national system rarely runs out. Shelves are almost always full, and a shopper in a well-served suburb has to try hard to encounter an empty aisle. Scarcity in America is less a question of whether stores have supplies than whether those supplies remain affordable and accessible. According to the USDA’s own annual report, food insecurity affected roughly 13.5 percent of American households in 2023, on the order of eighteen million households, up from around 12.8 percent the year before. That is a large minority for whom the question of dinner is not rhetorical, even as the national system distributes plenty. The two facts sit together: the system is abundant, and a substantial share of households struggle to afford it.
For the argument, that distinction matters. A household already stretching to afford food has no buffer to absorb a price increase, so the same trend that a comfortable household experiences as mild inflation registers for them as a tightening. Food-at-home prices rose by roughly a quarter between 2019 and 2023, one of the sharpest cumulative increases in decades, and while wages eventually caught up in the aggregate, the catch-up was uneven, and the shock landed first on those least able to absorb it.
Food deserts add a spatial dimension. The USDA has defined food deserts as areas with limited access to affordable, nutritious food, and by its 2017 count, some six thousand census tracts qualified, affecting roughly nineteen million people. The urban kind tends to lack supermarkets within walking distance; the rural kind tends to lack them within a reasonable drive. A national system that delivers to warehouses does not automatically deliver to a neighborhood without a store.
Beyond price and access, the earliest visible symptom of a thinning system is shrinking choice. Aisles do not go bare, but brands evaporate, products get swapped for cheaper stand-ins, and seasonal favorites become permanent absences. Organic options blink out for stretches. Cuts of meat contract toward the standard. Portions quietly shrink, the phenomenon often called shrinkflation, where the package gets smaller while the price holds or rises. Shoppers adapt by substituting and absorbing the cost, and the adjustment is quiet enough that the range of what is available can shrink in real time without anyone quite noticing.
Behind the shelf, the same logic runs at the level of the field: growers plant fewer varieties to hedge risk, softer produce crowds out labor-hungry kinds, and small diverse farms fold into larger, blander ones. Each choice is sensible on its own. Together, they strip out the diversity that resilience is built from.
The adaptation itself is the trap. People are remarkably flexible, and in a system like this, that flexibility doubles as a blindfold. Americans will not read thinning as scarcity. They will drift, shifting diets around what is on hand, accepting smaller portions or unfamiliar cuts, trading fresh food for processed stand-ins, and normalizing higher prices until they feel like weather. What once would have rung an alarm now shows up as a mild annoyance. By the time the public recognizes the system as thinner and more brittle, much of the damage will already be set, and reversing it will take longer than anyone wants to admit.
This is where the argument becomes genuinely uncomfortable, because the adaptiveness that protects people in the short run is also what allows the erosion to go unchallenged. A country that panicked at every shortage would force fixes; a country that shrugs and substitutes will tolerate a long slide. Every individual adjustment is reasonable, which is exactly why no single person ever feels the moment the baseline drops.
There is a deeper point about expectations. What people call normal is not a fixed standard but whatever they last got used to. A shopper in 2025 whose baseline was set in 2020 will experience a narrower, pricier store as normal rather than degraded, because they never knew the wider version. This ratchet runs in one direction: baselines rarely reset upward. Once the variety is gone and the price is absorbed, the lower level becomes the reference against which the next reduction is measured, and each generation inherits a slightly thinner normal than the last.
The trouble is never any one pressure. It is that they run in the same season, through the same narrow channels, and the risks do not so much add as amplify one another when they arrive together. A grower can usually survive a dry year, a distributor can usually survive a fuel spike, and a retailer can usually survive a labor gap, but even alone a severe drought or a major processing disruption can hurt, and when several arrive at once the system runs out of room to absorb them one at a time, having already spent its cushion on the first.
The 2020 meat disruption is the cleanest example, because it combined a health shock, a labor shock, and a concentration shock, and produced effects that no single cause would have generated. The 2021 Texas freeze is a second, combining weather, energy, and logistics. Neither was a food crisis in the sense of empty stores nationwide, but both showed how quickly a local event can travel when the network is thin.
Some places are more exposed than others, and the list is not exhaustive. The Central Valley leans on water-hungry crops and a shrinking aquifer, with a hard 2040 groundwater deadline approaching. The Ogallala plains depend on irrigation that is slowly running down. The Midwest answers to drought and tired soil. The South faces hurricanes and heat that punishes livestock. Cities depend on long supply chains with almost no slack built in.
An irony sits here worth stating plainly. The most exposure is now carried by regions that once fed themselves locally, while cities stay chained to a central system that keeps growing less sturdy. The two fail in opposite ways. Cities lose access first, producing almost nothing of their own and relying entirely on trucks and rail that can be delayed. Rural areas lose capacity first, because their land, water, and labor are the resources the system draws down, and the drawdown shows up locally long before it reaches a national statistic. Neither can bail out the other, since both draw from the same machinery.
Rather than repeat the whole catalogue, four things are worth holding in mind, because they are where the pressures actually meet:
The point of gathering them is not to list ten separate troubles but to show one system with several places it can give.
Here the ground shifts from measurement to scenario, and it is important to say so. Nothing in this section is a forecast in the strict sense. It is what could happen if the present trends continue and little is done to reverse them. Different assumptions produce different outcomes, and a determined policy response could change much of it.
Take the nearer term first. If several stressors converge within a few years, whether a prolonged drought in the West, a heat wave in the South, a fuel spike, a labor shortfall, or a plant failure or two, the plausible result is not collapse but a squeeze. Shelves stay populated, but with less variety. Seasonal fruits and vegetables get replaced by processed alternatives. Meat narrows toward standard cuts while specialty items thin out. Prices rise steadily over months rather than in a single shock, reshaping consumption without a moment anyone can point to. Restaurants simplify menus, and food deserts, already present, could widen if distribution costs push marginal stores out.
None of that is certain, and the word likely should be used carefully. The reduction in variety is plausible because the mechanisms behind it are already visible. The expansion of food deserts is a reasonable expectation but not a demonstrated trend. The disappearance of specialty items is a guess. What can be said with more confidence is that the system will keep running while quality, variety, and reliability slip a notch, and that by the end of such a stretch Americans will have adapted without protest and will believe the changes ordinary.
Extend the line to 2040 and the picture sharpens without turning catastrophic, though the sharpening depends on the assumptions. The 2040 date is not arbitrary: it is the deadline California set for its most stressed groundwater basins, so a real, measurable contraction of irrigated acreage in the Central Valley is built into current law unless the law changes. On that basis, and on the continuation of the other trends, a scenario for 2040 might include stores that remain open but with narrower choice, seasonal produce that is largely imported or industrially preserved, meat consumption narrowed to standard cuts, prices that keep climbing in small steps, and rural areas facing deeper water scarcity and poorer soil. Urban populations would lean harder on processed food, and consolidation would continue.
The honest way to present this is as one scenario among several, not as a prediction. A faster transition to drought-tolerant crops, a serious national push on storage and local processing, or a technological surprise could all soften the picture. A run of bad years or a major geopolitical shock could sharpen it. What the scenario does is show the direction of travel if nothing changes, which is the useful thing to know.
What emerges, if the scenario holds, is not collapse but persistent contraction, in which abundance gives way to sufficiency, reliability to uncertainty, and variety to standardization. None of that is dramatic, and that is the point. A country can lose resilience for twenty years and feel almost nothing, because the loss shows up as a slightly worse version of normal rather than as a break.
The American food supply is not going to disappear. It is becoming less forgiving, however, and its weaknesses run together: concentration without redundancy, aging growers and vanishing knowledge, harvest labor on a knife’s edge, rising input costs, a more volatile climate, water running down, soil running out, freight that must never stall, and a public with little reserve. None of this points to a sudden rupture. The future of food in this country will more likely be written by slow subtraction than by a single event, if it is written by drift at all.
Most of the work would be expensive and slow, and the honest thing is to say so. Rebuilt slack in the form of storage and reserves would help, but storage costs money every year and pays off only in bad ones, which is precisely why it keeps getting cut. Restored local capacity, meaning regional processing, regional distribution, and more food grown closer to where it is eaten, would shorten chains and reduce the number of things that can go wrong, but it fights economies of scale and would raise some prices. Rotation, cover crops, and reduced tillage would protect the soil base, but they demand patience and sometimes lower short-term yields.
Policy can help, and it can also mislead. Subsidies exist and are large, but they have historically favored a handful of commodity crops and large operations, which shapes what gets planted and who survives. Crop insurance is valuable, since it softens the blow of a bad year, but it can also encourage planting in marginal places. Environmental rules get enforced unevenly, and water rights are politically contested, so allocation follows power as much as hydrology. The gaps are not secret; they are hard to close, because each lever sits with a different set of hands, and none of them coordinates with the others.
Technology can blunt some risk, and it would be wrong to dismiss it. Precision agriculture, drought-tolerant seed, better irrigation, and improved cold chain all help, and none of it is free. Capital, know-how, and infrastructure are needed, and smaller producers often cannot reach them, so the benefits pool where the money already is. Technology also cannot fix a social problem: no sensor restores a retired farmer’s knowledge, and no app replaces a harvest crew that did not arrive.
What would make the biggest difference is probably the least glamorous thing, which is treating resilience as something worth paying for in advance rather than regretting in hindsight. That means valuing the buffer, the spare plant, the local mill, the cover crop, and the extra warehouse, even when they look inefficient in a good year. It means measuring the system not by how cheaply it runs in calm weather but by how gracefully it fails in bad weather.
The hardest part is that resilience has no constituency. It has no ribbon-cutting, no dramatic rescue, no visible hero. A storage facility that quietly prevents a shortage is never celebrated, and a plant that stays open is never news. The rewards of redundancy are invisible precisely when they work, which is why they keep losing to the visible rewards of cutting costs. Changing that would require a shift in how the system measures success, and such shifts are slow.
The alternative is to keep adapting, keep shrugging, and keep calling each new reduction normal. That path is not catastrophic. It is a long, quiet slide, punctuated by a few bad years that everyone experiences as bad luck rather than as the bill coming due.
What makes the situation worth watching is not the drama it promises but the drama it withholds. The shelves stay full, the lights stay on, the strawberries keep appearing in February, and the system keeps running right up until the day it runs a little worse than the day before. By then the adjustment will already be underway, and most people will call it nothing at all.
There is no day when the country wakes up to a food crisis. There is only a sequence of ordinary weeks in which the margin gets a little thinner each time, and a population that adapts so gracefully it never quite notices what it has lost. The food is still there. The ease of having it is what is quietly going away.
If there is a single idea worth carrying out of all this, it is that abundance and resilience are different things, and the country has been measuring only the first. Abundance is what you can see on the shelf. Resilience is the room behind it, the part that only shows up in a bad year. For a long time the two moved together, which is why nobody learned to tell them apart. They have started to separate, and the separation is the whole story. What matters now is whether anyone notices while there is still time to close the gap, or whether the country simply adjusts, again and again, and calls the result normal.
Tweet
Continue reading...
[ H/T The Burning Platform ]
Just after the morning deliveries and before the first customers, a supermarket looks both ordinary and slightly deceptive. The lights are on, the coolers hum, and the shelves hold exactly what they held last week. Strawberries show up in February, lettuce comes from ground that was desert two generations ago, and on the same shelf beef raised across three states sits beside bread baked from grain cut a thousand miles away. Someone pushing a cart through the whole store can finish the errand without once wondering how any of it got there.
That is part of why the American food supply is so easy to take for granted. Nobody praises the trucking fleets, the cold storage, the seed suppliers, or the packing houses, because a thing that rarely fails stops being noticed. Two or three generations have grown up treating food the way they treat running water: as a fact of the world, unremarkable and about as interesting.
Reliability, though, is not the same as resilience, and the two are easy to confuse because the difference only appears when something goes wrong. A farm can still bring in a decent harvest during a dry year, until that dry year coincides with expensive fertilizer, a thin labor crew, and a delayed shipment. Shelves can stay full while carrying fewer brands at higher prices. The country is not running out of food. What it may be running out of is the cushion that once let a bad season pass without anyone noticing.
This article rests on three kinds of claims, and it is worth keeping them apart. First come the demonstrated problems: measured facts about the age of farmers, the drawdown of aquifers, the concentration of meatpacking, the cost of inputs, the price of food. Then comes an interpretation, which is mine and can be argued with: the suggestion that these separate trends, taken together, are eroding the system’s ability to absorb a shock. Last come scenarios, which are conditional rather than predictive, describing what could happen over the next decade or two if several pressures arrive at once and little changes. The text will say which is which as it goes.
What follows is not a forecast of collapse. It is an attempt to look past the grocery aisle at the people who grow the food, the water and soil they depend on, the infrastructure that moves it, and the habits of a public that has rarely had a reason to worry.
A system built to keep moving
American food logistics are organized around movement rather than storage. Warehouses are designed for turnover, distribution centers run on delivery windows that assume trucks arrive when promised, and most supermarkets carry only a few days of stock, though the exact figure varies by product; fresh produce and bakery items turn over within days, while canned and frozen goods sit much longer. Very little slack exists by design, because idle inventory costs money and motion does not.
This is what the industry calls just-in-time, and for decades it worked remarkably well. Waste fell, prices stayed low, and a range of food reached the shelf that no earlier era could match. The bill for that efficiency comes due only under stress, and when it does, the buffers get spent quickly. A single failed truck route can empty a produce section for days, and one plant going dark for repairs can thin meat counters across several states. Fewer alternate paths run through the network than most people assume, and the paths that do exist tend to funnel through the same handful of junctions.
Older food economies leaned on redundancy instead. Grain sat in silos, towns fed themselves, and a surplus worked as insurance against a bad year. In the modern version, surplus reads as waste and storage reads as cost, so redundancy survives mainly where it happens to escape the spreadsheet. What remains runs beautifully when conditions hold and begins to creak the moment several things break at once. The efficiency is real, and so is the fragility that comes with it.
Concentration belongs to the same design. A small number of plants handles an enormous share of the nation’s meat, grain, and packaged food. Four companies — Cargill, JBS, Tyson, and National Beef — slaughter roughly 80 to 85 percent of the country’s fed cattle, a figure the Government Accountability Office put at about 85 percent in a 2021 report and which has moved little since. The top four pork firms account for roughly two-thirds of hog slaughter, and poultry is more concentrated still. On the grain side, four international traders, often nicknamed the ABCD firms, have long handled a large share of world flows, with American exports passing through a shrinking number of elevators and ports.
When the packing sector sneezes, the meat case in every region catches cold. The clearest recent example came in the spring of 2020, when COVID-19 tore through crowded slaughterhouses. A Smithfield pork plant in Sioux Falls, South Dakota, processing something like 4 to 5 percent of the nation’s pork at the time, shut down after becoming one of the largest single-source outbreaks in the country. A Tyson beef plant in Columbus Junction, Iowa, closed. JBS and National Beef idled facilities in Colorado and Kansas. At the peak, beef and pork production fell by roughly a quarter, and the shock moved in both directions at once: prices paid to farmers collapsed, sometimes to the point that hogs were euthanized for lack of processing capacity, while retail meat prices climbed. The same disruption reached producers and consumers from opposite sides, which is what concentration does. It transmits a local failure across the whole chain.
None of this makes consolidation irrational. Real reasons stood behind it: economies of scale, thin margins, decades of mergers, and a regulatory climate that looked kindly on size. A large plant can slaughter more animals per hour at lower cost than a small one, and consumers have benefited through cheaper meat. What the system traded away was its detours, and with them the ability to route around a failure.
Leverage hides inside that structure without anyone seizing it deliberately. It emerges on its own from squeezing production and transport into fewer hands in the name of cost. Abundance becomes conditional rather than guaranteed, and a small number of failures can move food away from millions of tables at once.
Interpretation begins here. It is tempting to say that just-in-time always made the system fragile, but that oversimplifies. Grocery stores have always carried limited shelf stock; a butcher in the 1950s did not keep weeks of beef either. What has changed is the network behind the shelf, with fewer plants, fewer carriers, fewer alternate routes, and thinner reserves at every layer. The modern system is not more fragile because it moves goods. It is more fragile because the number of places where a single failure can stop the flow has grown while the number of ways around those places has shrunk.
Rebuilding redundancy is slow. Storage, local processing, and alternative routes take years to restore once they have been allowed to disappear, and the disappearance happens on a schedule of cost savings rather than of visible shortages, so the consequences arrive long after the decisions that caused them. A regional plant closed in 2015 to save money does not show its cost until a disruption in 2025, by which point the plant is a parking lot.
One wrinkle rarely enters the discussion. Efficiency and resilience are not opposites so much as competing goods, and the trade-off stays invisible because its cost is deferred. Every dollar saved by running leaner shows up immediately on a balance sheet, while every unit of resilience given up materializes only in a bad year. In good years the lean system looks strictly superior, which is why it keeps winning and why the arguments for holding reserve keep losing. The logic is not wrong. It is incomplete, because it prices the present and discounts the future.
That asymmetry explains a great deal about how the system got where it is. Managers are rewarded for the savings they can demonstrate and only occasionally punished for the resilience they quietly surrender. A warehouse that keeps a little extra stock looks wasteful until the day a truck is late, and by then the person who approved the extra stock may have moved on. Over decades these small decisions compound in one direction, and the system ends up leaner than anyone consciously chose.
The same logic runs through transportation. Trucking moves the large majority of American freight by weight, commonly estimated at around 70 percent, and the industry has cycled through driver shortages for years, with trade associations citing gaps in the tens of thousands. Rail, which hauls grain, fertilizer, and other bulk goods, has consolidated down to a handful of Class I carriers, and service complaints spiked in 2022 as the railroads cut crews and lengthened trains in pursuit of their own efficiency. Ports handle imports of everything from coffee to out-of-season produce, and when they clog, as Los Angeles and Long Beach did through 2021, the delay propagates through the whole country.
None of these links is broken. Each is simply running with less give than it once had, and give is exactly what a shock consumes.
The people and resources behind the harvest
The hands working the land are older than they used to be, a pattern any drive through farm country will show and the numbers confirm. According to the 2022 Census of Agriculture, published by the USDA’s National Agricultural Statistics Service, the average age of an American farm producer has reached 58.1 years, and the share past 65 keeps rising. Fewer young people are stepping in behind them.
The reasons are not mysterious. Farming means heavy debt, expensive equipment, volatile input prices, and weather that no longer keeps its word, and a single ruined season can erase years of work. Against a steady city paycheck, that is a hard case to make to someone choosing a career. So when growers retire, the land often fails to pass to new hands. It gets bought by larger operations, cut into subdivisions, leased for something other than food, or simply left idle because keeping it in production no longer pays.
The scale of the shift is easy to understate. The 2022 census counted about 1.9 million farms, down from roughly 6.8 million in 1935, and the long decline has never really reversed. Land in farms, a category that includes pasture and range as well as cropland, totaled about 880 million acres in 2022, down slightly from 2017. Cropland proper is smaller, on the order of 300 million acres, and it matters not to blur the two, since a ranch in Nevada and a cornfield in Iowa are counted under the same headline even though they function very differently. What matters for the food supply is the cropland, and the trend there is gentler than the farm-count trend, though still downward at the edges.
A second pattern sits underneath the averages. Most farms are small, and most production comes from a few of them. Roughly 9 percent of farms, those with sales above one million dollars, account for the bulk of the total value of what American agriculture produces, while the bottom half of farms by sales contribute only a sliver. The median farm household often earns more from off-farm work than from farming itself. This is less a sign of failure than a structural fact: the business of growing food has concentrated into fewer, larger, more capitalized operations, while the small farm survives as a part-time pursuit.
The consequence for resilience is not obvious. A large operation can invest in irrigation, storage, and equipment that a small one cannot afford, which arguably makes it more robust. But the same operation is also more exposed to a single shock, whether a bad commodity price, a failed crop, or an interest-rate move on a large loan, and the loss of many small, diversified farms removes a kind of distributed redundancy that the system once had without anyone counting it.
A second loss draws even less attention, and it is buried inside that shift: knowledge leaves with the people who hold it. When to plant a difficult crop, how to read a particular patch of soil, which pests arrive in which week, how to carry a marginal field through a dry August. None of that lives in a manual. It passes down through generations and goes quiet when a farm changes hands. Machinery can replace some of it and not the rest, and the part it cannot replace is often the part that matters most when conditions turn strange.
The coming handover is substantial. Farmland is old, and a great deal of it will change hands in the next couple of decades as the current generation retires. Estimates vary, but the American Farmland Trust has projected that on the order of several hundred million acres could transition ownership in the next twenty years, and whether that land passes to working farmers or to developers and investors will shape what the countryside looks like. When a farm is sold to a larger neighbor, the land stays in production but the family knowledge often does not transfer with it; when it is sold for houses, the production stops altogether.
Small fruit growers in Washington state offer one illustration. Orchards there have long depended on skilled hand-picking, and older growers perfected a judgment about ripeness, timing, and handling that no machine has fully matched. As those growers retire, the expertise thins, and yields can suffer even though the land itself is unchanged.
Labor enters the picture next. Much of American agriculture runs on seasonal work, and seasonal work answers to biology rather than to anyone’s schedule. Fruit and vegetables ripen inside narrow windows, and missing the window loses the crop no matter how good the growing season was. Labor gaps have already left produce rotting in fields for want of pickers. The work leans heavily on immigrant crews, and a large share of hired farmworkers are foreign-born, with a substantial minority lacking legal status. The estimates vary, but the foreign-born share is well above half in many crops. Any kink in that pipeline turns directly into food left to spoil in the sun.
The H-2A visa program, which brings in temporary agricultural workers, has grown dramatically as a result, from roughly eighty thousand certified positions a little over a decade ago to several hundred thousand today. That growth is itself a signal: American agriculture increasingly cannot staff its harvests from the domestic labor pool alone, and the program that fills the gap is subject to political weather.
Timing is what makes the labor question brittle. A factory can pause a line and restart it later without losing much. A field cannot. Crops do not wait for a policy debate to settle and do not care what the calendar says. A shortage does not simply reduce the harvest; it reduces the harvest at the exact moment the crop is worth the most, because fruit that was perfect on Monday is mush by Friday. Processing plants that expected a delivery sit idle, freight schedules slip, and availability drops just as demand peaks.
Tomatoes must be picked and moved within a short span before they soften. Berries tolerate almost no delay and almost no rough handling. Lettuce and leafy greens are cut in tight windows and spoil quickly. Each depends on crews arriving exactly when they are needed, so any disruption, whether a change in immigration policy, an economic downturn, or a public health emergency, lands directly on harvest timing.
On top of labor sits cost. Modern growing depends on a web of inputs most people never picture: diesel for machines and freight, fertilizer produced through energy-hungry chemistry, spare parts drawn through global supply chains, irrigation tied to water that keeps moving, and packaging hostage to industrial cycles. Fertilizer prices in particular spiked in 2021 and 2022, in some cases doubling or tripling, driven by natural gas costs and the war in Ukraine, since nitrogen fertilizer is essentially made from natural gas. Diesel followed oil, and farm debt has climbed past half a trillion dollars.
When those costs surge, farmers do not usually fold right away. They get careful. They plant fewer acres, walk away from marginal ground, choose crops that demand less labor and carry less risk, and cut output to protect the balance sheet. Every decision is sensible from where the grower stands, and every decision narrows the national supply a little further.
For a customer, the shrinkage stays invisible for a season or two. It accumulates quietly, spread across regions and years, until the gap reaches the shelf and nobody can say where it came from. Four adjustments recur: fewer acres planted, marginal land abandoned, a shift toward low-risk crops with steady demand, and lower overall output to stay solvent. Each move is rational on its own, and collectively they shave both production and resilience at the same time. No farmer sets out to weaken the food supply. The weakening falls out of thousands of private decisions made for private reasons, none of them coordinated, all of them pointing the same way when costs keep rising.
Consumers rarely see one piece of this, and it deserves its own moment because it reframes who pays. The share of the retail food dollar that reaches the farmer is small and has been small for a long time, on the order of fifteen cents by the USDA’s own accounting, with the rest absorbed by processing, packaging, transport, marketing, and retail. When retail prices rise, the increase does not flow evenly back to the farm. During the 2020 meat disruption, the gap between what farmers received and what shoppers paid widened sharply, and much of the spread stayed with the middle of the chain even after production recovered. The point is not that anyone is cheating. It is that the farm is a small and vulnerable node within a long chain, and shocks that look like a food problem from the aisle often register at the farm as a price collapse.
Water, soil, and weather
Some of the pressure comes from the ground and the sky, and the mechanisms are worth separating, because they are often lumped together into a vague sense of climate without anyone distinguishing what is actually happening.
Start with water. The Ogallala Aquifer, which underlies parts of eight plains states from South Dakota to Texas, supplies a significant share of American irrigation, often cited at around 30 percent of the nation’s irrigation groundwater, and it has been declining for decades. The decline is uneven: some areas have dropped by more than a hundred feet since large-scale pumping began, while others have been steadier. Kansas, which depends heavily on the aquifer, has recorded some of the steepest losses, and wells have run dry in parts of the Texas Panhandle. The depletion is patient rather than dramatic. Where water becomes uncertain, planting becomes a bet, and land without reliable water cannot reliably feed anyone. Over-extraction leaves its own damage: sinking ground, rising salinity, falling yields.
California’s Central Valley tells a related story with different mechanics. The valley is the core of a state that grows a large share of the nation’s vegetables and tree fruit, and it leans on both surface water and groundwater. During droughts, farmers pump harder, and the aquifer beneath the valley has dropped in some places dramatically, causing the ground itself to sink, with subsidence that has damaged canals and roads. The state’s Sustainable Groundwater Management Act, passed in 2014, requires the most stressed basins to reach sustainability by 2040, which sets up a genuine reckoning: bringing pumping into balance will likely mean taking a substantial amount of land out of irrigated production. That 2040 deadline is one of the few hard dates in this story, and it is worth remembering when the later scenarios arrive.
The Colorado River adds a third case. The river’s flow is allocated among seven states and Mexico under a compact written in 1922, when the river was running higher than it does today, and the allocations exceed the actual supply in most years. A long drought, punctuated by negotiations and emergency cuts, has forced the basin’s users to reduce withdrawals, and the lower basin states have committed to significant temporary reductions. The river irrigates a great deal of winter vegetable production in Arizona and California, so what happens on the Colorado reaches dinner plates across the country.
Soil comes next, and the common shorthand that chemical fertilizer erodes fertility needs a little care. The heavier problem is more physical than chemical. Decades of intensive tillage have stripped organic matter and broken up soil structure, and degraded soil holds less water, resists erosion less well, and recovers more slowly. Fertilizer can mask some of that loss by supplying nutrients directly, which is why yields have stayed high even as the soil underneath has thinned. It is a patch rather than a repair, and ground that has been mined rather than farmed does not recover on its own. Rebuilding it takes cover crops, rotation, and patience, the kind of patient work a system tuned to quarterly results does not always reward.
The numbers here are hard to pin precisely, but the direction is well established. Soil erosion on American cropland runs to roughly one and a half to two billion tons a year by common estimates, most of it tolerable in any given year but cumulative over decades. Cover crops, one of the main remedies, are used on a small minority of cropland, on the order of five percent, though growing. No-till and reduced tillage are more widespread but uneven. The farmer and writer Wendell Berry called soil “the great connector of lives, the source and destination of all,” and the line reads less like poetry and more like a specification once you understand what healthy ground actually does.
The weather itself comes last. Farmers once leaned on patterns inherited across generations, reading rainfall, planting windows, and the rough shape of a normal season from what their parents and grandparents had seen. That inheritance is coming apart, though coming apart is itself a metaphor, and the concrete changes are more useful. From 2000 into the early 2020s, southwestern North America experienced a megadrought that researchers, in a widely cited 2022 study in Nature Climate Change, judged the worst in at least twelve hundred years. Global average temperatures set records in 2023 and again in 2024. The USDA redrew its plant hardiness zone map in 2023, and much of the country shifted warmer. Flash droughts, sudden and intense, have become more common, and the extremes have grown more extreme, with heavy downpours in some places, prolonged heat in others, and the same region sometimes seeing both within a year.
Gradual change is manageable. Volatility is harder, because when the pattern stops being trustworthy, planning turns into guesswork. A grower can adapt to a warmer average by planting earlier or switching varieties. A grower cannot easily adapt to a season that swings from flood to drought and back, because the equipment, the contracts, and the calendar are all built around a rhythm that no longer holds. Heat that once came in predictable waves now lands at odd hours, stressing animals and trimming yields in ways growers struggle to anticipate.
The 2021 Texas freeze, Winter Storm Uri, offered a preview of how these pressures interact. The storm knocked out power and water across the state, froze citrus and other crops, and disrupted food processing and transport, all at once. It was a weather event, but its damage ran through the same channels that every other pressure uses: energy, water, labor, and logistics.
Each factor alone would be survivable. Together they compound, and a system never designed for persistent unpredictability feels the strain in every decision a grower makes. It is also worth noting that climate is not a single lever. A warmer average, a drier aquifer, and a more volatile season are three different problems with three different remedies, and conflating them makes the response vaguer than it needs to be.
The distance between farm and table
Food in this country routinely crosses thousands of miles between field and fork, which requires cheap fuel, available drivers, working highways and rail, and ports that keep flowing. A snag anywhere delays everything downstream. Often the food exists and simply cannot arrive in time, a distinction that matters more than it sounds, because shortages frequently erupt not because nothing was grown but because what was grown could not move. A household staring at a bare shelf has no way to tell the difference.
The idea of food miles is often overstated, since a tomato trucked from Mexico can carry a smaller carbon footprint than one grown in a heated local greenhouse, but distance matters for a different reason: it multiplies the number of things that can go wrong. Every additional mile is another chance for a breakdown, another border, another handoff. A short local chain has fewer failure points by construction, even if each one is not inherently more reliable.
Four weak points carry most of the risk. Fuel spikes make every mile costlier and push marginal routes toward unprofitability. Driver shortages limit how much can be delivered even when the trucks exist, and the trucking industry has cycled through shortages for years, with the American Trucking Associations estimating gaps in the tens of thousands and warning they could grow. Port congestion stalls the imports and exports the domestic system depends on. And aging infrastructure, from bridges to rail lines to terminals older than most of the people using them, raises the odds of a breakdown along a route with no realistic alternate.
The 2021 port crisis is instructive. Backed-up container ships off Los Angeles and Long Beach, driven by a surge in imports and pandemic-era bottlenecks, delayed everything from furniture to coffee and forced retailers to reroute and ration. The ports eventually cleared, but the episode showed how a single chokepoint on the coast can ripple into shelves two thousand miles inland. Rail had its own version of the same problem in 2022, when service deteriorated badly enough that regulators held hearings and shippers complained of missed deliveries and unpredictable schedules.
Cold chain adds a layer most people never consider. Produce, meat, dairy, and frozen goods all depend on an unbroken sequence of refrigerated storage and transport, and a break anywhere, whether a power outage at a warehouse or a failed compressor on a truck, can spoil a whole load. The cold chain is one of the genuine engineering achievements of the modern food system and one of the most brittle, because it has no tolerance for interruption.
The infrastructure itself is old. A large share of American bridges are past their design life, and the rail network, while privately owned and generally well maintained on the main lines, has seen deferred maintenance on the margins. The country has underinvested in the physical layer of its food system for decades, not because anyone decided to, but because maintenance is invisible and unglamorous while new projects are visible and political.
Fuel deserves a special note, because it sits underneath everything else. Diesel powers the trucks, the trains, the tractors, and the ships that move food, and it also powers the fertilizer and the irrigation pumps. When diesel prices spike, as they did in 2022, the cost propagates through every link at once, and it hits the most marginal operators hardest, meaning the small trucking firms, the family farms, and the regional distributors, which is to say exactly the nodes that provide redundancy to the system. A fuel spike does not just make food cost more. It quietly removes some of the smaller players who would otherwise have absorbed a shock.
None of these needs to fail completely to matter. Each only has to slow down, and the delays stack. A driver shortage, a port backup, and a fuel spike in the same quarter do not add up to three problems. They multiply into one much larger one, because the system has no slack to absorb them in sequence.
What consumers notice first
The softest vulnerability may live inside people’s heads, but it is worth being precise about what the public actually experiences, because the easy claim that Americans have never known scarcity is only partly true and can mislead.
It is true that the national system rarely runs out. Shelves are almost always full, and a shopper in a well-served suburb has to try hard to encounter an empty aisle. Scarcity in America is less a question of whether stores have supplies than whether those supplies remain affordable and accessible. According to the USDA’s own annual report, food insecurity affected roughly 13.5 percent of American households in 2023, on the order of eighteen million households, up from around 12.8 percent the year before. That is a large minority for whom the question of dinner is not rhetorical, even as the national system distributes plenty. The two facts sit together: the system is abundant, and a substantial share of households struggle to afford it.
For the argument, that distinction matters. A household already stretching to afford food has no buffer to absorb a price increase, so the same trend that a comfortable household experiences as mild inflation registers for them as a tightening. Food-at-home prices rose by roughly a quarter between 2019 and 2023, one of the sharpest cumulative increases in decades, and while wages eventually caught up in the aggregate, the catch-up was uneven, and the shock landed first on those least able to absorb it.
Food deserts add a spatial dimension. The USDA has defined food deserts as areas with limited access to affordable, nutritious food, and by its 2017 count, some six thousand census tracts qualified, affecting roughly nineteen million people. The urban kind tends to lack supermarkets within walking distance; the rural kind tends to lack them within a reasonable drive. A national system that delivers to warehouses does not automatically deliver to a neighborhood without a store.
Beyond price and access, the earliest visible symptom of a thinning system is shrinking choice. Aisles do not go bare, but brands evaporate, products get swapped for cheaper stand-ins, and seasonal favorites become permanent absences. Organic options blink out for stretches. Cuts of meat contract toward the standard. Portions quietly shrink, the phenomenon often called shrinkflation, where the package gets smaller while the price holds or rises. Shoppers adapt by substituting and absorbing the cost, and the adjustment is quiet enough that the range of what is available can shrink in real time without anyone quite noticing.
Behind the shelf, the same logic runs at the level of the field: growers plant fewer varieties to hedge risk, softer produce crowds out labor-hungry kinds, and small diverse farms fold into larger, blander ones. Each choice is sensible on its own. Together, they strip out the diversity that resilience is built from.
The adaptation itself is the trap. People are remarkably flexible, and in a system like this, that flexibility doubles as a blindfold. Americans will not read thinning as scarcity. They will drift, shifting diets around what is on hand, accepting smaller portions or unfamiliar cuts, trading fresh food for processed stand-ins, and normalizing higher prices until they feel like weather. What once would have rung an alarm now shows up as a mild annoyance. By the time the public recognizes the system as thinner and more brittle, much of the damage will already be set, and reversing it will take longer than anyone wants to admit.
This is where the argument becomes genuinely uncomfortable, because the adaptiveness that protects people in the short run is also what allows the erosion to go unchallenged. A country that panicked at every shortage would force fixes; a country that shrugs and substitutes will tolerate a long slide. Every individual adjustment is reasonable, which is exactly why no single person ever feels the moment the baseline drops.
There is a deeper point about expectations. What people call normal is not a fixed standard but whatever they last got used to. A shopper in 2025 whose baseline was set in 2020 will experience a narrower, pricier store as normal rather than degraded, because they never knew the wider version. This ratchet runs in one direction: baselines rarely reset upward. Once the variety is gone and the price is absorbed, the lower level becomes the reference against which the next reduction is measured, and each generation inherits a slightly thinner normal than the last.
When risks compound
The trouble is never any one pressure. It is that they run in the same season, through the same narrow channels, and the risks do not so much add as amplify one another when they arrive together. A grower can usually survive a dry year, a distributor can usually survive a fuel spike, and a retailer can usually survive a labor gap, but even alone a severe drought or a major processing disruption can hurt, and when several arrive at once the system runs out of room to absorb them one at a time, having already spent its cushion on the first.
The 2020 meat disruption is the cleanest example, because it combined a health shock, a labor shock, and a concentration shock, and produced effects that no single cause would have generated. The 2021 Texas freeze is a second, combining weather, energy, and logistics. Neither was a food crisis in the sense of empty stores nationwide, but both showed how quickly a local event can travel when the network is thin.
Some places are more exposed than others, and the list is not exhaustive. The Central Valley leans on water-hungry crops and a shrinking aquifer, with a hard 2040 groundwater deadline approaching. The Ogallala plains depend on irrigation that is slowly running down. The Midwest answers to drought and tired soil. The South faces hurricanes and heat that punishes livestock. Cities depend on long supply chains with almost no slack built in.
An irony sits here worth stating plainly. The most exposure is now carried by regions that once fed themselves locally, while cities stay chained to a central system that keeps growing less sturdy. The two fail in opposite ways. Cities lose access first, producing almost nothing of their own and relying entirely on trucks and rail that can be delayed. Rural areas lose capacity first, because their land, water, and labor are the resources the system draws down, and the drawdown shows up locally long before it reaches a national statistic. Neither can bail out the other, since both draw from the same machinery.
Rather than repeat the whole catalogue, four things are worth holding in mind, because they are where the pressures actually meet:
- The system has very little stored reserve, so a delay anywhere turns into a shortage somewhere else.
- The most critical nodes, from packing plants to ports, are few enough that one failure travels nationally.
- The natural base, meaning water and soil, is degrading slowly and unevenly, with the damage concentrated in the regions that produce the most.
- The public has little reserve and adapts quickly, which delays any political pressure to fix the problem.
The point of gathering them is not to list ten separate troubles but to show one system with several places it can give.
What the next decades might bring
Here the ground shifts from measurement to scenario, and it is important to say so. Nothing in this section is a forecast in the strict sense. It is what could happen if the present trends continue and little is done to reverse them. Different assumptions produce different outcomes, and a determined policy response could change much of it.
Take the nearer term first. If several stressors converge within a few years, whether a prolonged drought in the West, a heat wave in the South, a fuel spike, a labor shortfall, or a plant failure or two, the plausible result is not collapse but a squeeze. Shelves stay populated, but with less variety. Seasonal fruits and vegetables get replaced by processed alternatives. Meat narrows toward standard cuts while specialty items thin out. Prices rise steadily over months rather than in a single shock, reshaping consumption without a moment anyone can point to. Restaurants simplify menus, and food deserts, already present, could widen if distribution costs push marginal stores out.
None of that is certain, and the word likely should be used carefully. The reduction in variety is plausible because the mechanisms behind it are already visible. The expansion of food deserts is a reasonable expectation but not a demonstrated trend. The disappearance of specialty items is a guess. What can be said with more confidence is that the system will keep running while quality, variety, and reliability slip a notch, and that by the end of such a stretch Americans will have adapted without protest and will believe the changes ordinary.
Extend the line to 2040 and the picture sharpens without turning catastrophic, though the sharpening depends on the assumptions. The 2040 date is not arbitrary: it is the deadline California set for its most stressed groundwater basins, so a real, measurable contraction of irrigated acreage in the Central Valley is built into current law unless the law changes. On that basis, and on the continuation of the other trends, a scenario for 2040 might include stores that remain open but with narrower choice, seasonal produce that is largely imported or industrially preserved, meat consumption narrowed to standard cuts, prices that keep climbing in small steps, and rural areas facing deeper water scarcity and poorer soil. Urban populations would lean harder on processed food, and consolidation would continue.
The honest way to present this is as one scenario among several, not as a prediction. A faster transition to drought-tolerant crops, a serious national push on storage and local processing, or a technological surprise could all soften the picture. A run of bad years or a major geopolitical shock could sharpen it. What the scenario does is show the direction of travel if nothing changes, which is the useful thing to know.
What emerges, if the scenario holds, is not collapse but persistent contraction, in which abundance gives way to sufficiency, reliability to uncertainty, and variety to standardization. None of that is dramatic, and that is the point. A country can lose resilience for twenty years and feel almost nothing, because the loss shows up as a slightly worse version of normal rather than as a break.
Building resilience
The American food supply is not going to disappear. It is becoming less forgiving, however, and its weaknesses run together: concentration without redundancy, aging growers and vanishing knowledge, harvest labor on a knife’s edge, rising input costs, a more volatile climate, water running down, soil running out, freight that must never stall, and a public with little reserve. None of this points to a sudden rupture. The future of food in this country will more likely be written by slow subtraction than by a single event, if it is written by drift at all.
Most of the work would be expensive and slow, and the honest thing is to say so. Rebuilt slack in the form of storage and reserves would help, but storage costs money every year and pays off only in bad ones, which is precisely why it keeps getting cut. Restored local capacity, meaning regional processing, regional distribution, and more food grown closer to where it is eaten, would shorten chains and reduce the number of things that can go wrong, but it fights economies of scale and would raise some prices. Rotation, cover crops, and reduced tillage would protect the soil base, but they demand patience and sometimes lower short-term yields.
Policy can help, and it can also mislead. Subsidies exist and are large, but they have historically favored a handful of commodity crops and large operations, which shapes what gets planted and who survives. Crop insurance is valuable, since it softens the blow of a bad year, but it can also encourage planting in marginal places. Environmental rules get enforced unevenly, and water rights are politically contested, so allocation follows power as much as hydrology. The gaps are not secret; they are hard to close, because each lever sits with a different set of hands, and none of them coordinates with the others.
Technology can blunt some risk, and it would be wrong to dismiss it. Precision agriculture, drought-tolerant seed, better irrigation, and improved cold chain all help, and none of it is free. Capital, know-how, and infrastructure are needed, and smaller producers often cannot reach them, so the benefits pool where the money already is. Technology also cannot fix a social problem: no sensor restores a retired farmer’s knowledge, and no app replaces a harvest crew that did not arrive.
What would make the biggest difference is probably the least glamorous thing, which is treating resilience as something worth paying for in advance rather than regretting in hindsight. That means valuing the buffer, the spare plant, the local mill, the cover crop, and the extra warehouse, even when they look inefficient in a good year. It means measuring the system not by how cheaply it runs in calm weather but by how gracefully it fails in bad weather.
The hardest part is that resilience has no constituency. It has no ribbon-cutting, no dramatic rescue, no visible hero. A storage facility that quietly prevents a shortage is never celebrated, and a plant that stays open is never news. The rewards of redundancy are invisible precisely when they work, which is why they keep losing to the visible rewards of cutting costs. Changing that would require a shift in how the system measures success, and such shifts are slow.
The alternative is to keep adapting, keep shrugging, and keep calling each new reduction normal. That path is not catastrophic. It is a long, quiet slide, punctuated by a few bad years that everyone experiences as bad luck rather than as the bill coming due.
The margin
What makes the situation worth watching is not the drama it promises but the drama it withholds. The shelves stay full, the lights stay on, the strawberries keep appearing in February, and the system keeps running right up until the day it runs a little worse than the day before. By then the adjustment will already be underway, and most people will call it nothing at all.
There is no day when the country wakes up to a food crisis. There is only a sequence of ordinary weeks in which the margin gets a little thinner each time, and a population that adapts so gracefully it never quite notices what it has lost. The food is still there. The ease of having it is what is quietly going away.
If there is a single idea worth carrying out of all this, it is that abundance and resilience are different things, and the country has been measuring only the first. Abundance is what you can see on the shelf. Resilience is the room behind it, the part that only shows up in a bad year. For a long time the two moved together, which is why nobody learned to tell them apart. They have started to separate, and the separation is the whole story. What matters now is whether anyone notices while there is still time to close the gap, or whether the country simply adjusts, again and again, and calls the result normal.
Tweet
Continue reading...
[ H/T The Burning Platform ]
