Stating the theory clearly
The 3 to 7 percent theory holds that when a tightly integrated agricultural region loses 3 to 7 percent of its productive farmland to non-agricultural uses, it crosses a threshold that triggers cascading economic harm far beyond the converted acres. The damage is not proportional. Losing 5 percent of the farmland does not mean losing 5 percent of the economy. It means destabilizing the feed supply, compressing dairy margins, pushing vendors toward closure thresholds, weakening processor throughput, displacing worker families, and eroding the tax base that funds schools and services. The ripple effect touches every sector because every sector is coupled to the land.
This is not a theory about one farm going out of business. It is a theory about what happens when a system built on interdependence loses a piece that everything else depends on. In a region like the Magic Valley, where agriculture accounts for 42 percent of all jobs and 59 percent of all sales according to the University of Idaho Extension's 2018 economic impact analysis1, the farmland is not just dirt. It is the foundation of a $20 billion regional economy.
The numbers: what is at stake
Before examining the cascade mechanism, the scale of what is at risk must be stated plainly with real data.
The agricultural economy. Idaho's dairy industry alone generated $3.9 billion in farm-gate receipts in 2025, according to Idaho Dairymen's Association data3. The direct economic impact is approximately $7 billion; with indirect effects, the total exceeds $11 billion statewide3. The industry supports 33,000 jobs and generates $155 million in state and local taxes3. Seventy to seventy-two percent of Idaho's dairy cows are concentrated in the Magic Valley. The region's 350 dairy operations produced 18.26 billion pounds of milk in 2025, narrowly edging Texas for the third-highest state production nationally3.
The land. The USDA Natural Resources Conservation Service classifies soils into capability classes. Class I and Class II soils are the nation's most productive agricultural land, capable of sustained intensive cultivation with minimal conservation practices. In Idaho's Magic Valley, much of the irrigated farmland falls into these prime classifications. The American Farmland Trust's "Farms Under Threat" analysis found that 83 percent of Idaho's projected farmland conversion will occur on the state's best agricultural land2. These are not marginal acres. They are among the most productive soils in the western United States, made viable by the irrigation infrastructure built over a century of public and private investment tied to the Snake River and the Eastern Snake Plain Aquifer.
The conversion rate. From 2017 to 2022, Idaho lost 144,000 acres of farmland, declining from 11.69 million to 11.55 million acres, according to Census of Agriculture data reported by the Idaho Farm Bureau Federation5. From 2016 to 2023, the state lost approximately 65,000 acres, equivalent to 400 farms, $40 million in agricultural output, and 900 jobs2. The American Farmland Trust found that the actual conversion rate is 96 times faster than previously projected2. Idaho is losing approximately 14 square miles of agricultural land per year5.
What is replacing it. Data centers and large-scale energy installations are consuming irrigated farmland at an accelerating pace. Meta's $800 million data center occupies 485 acres of former family farms and a beef processing operation near Kuna10. The Gemstone Technology Park's $1 billion data center was approved on 620 acres in 202511. In Jerome County, Cat Creek Energy has proposed a $120 million solar farm on 554 acres south of Eden10. Another proposal would install 150,000 solar panels across 405 acres10. These projects require the same inputs agriculture requires: open land, water, power, and manageable climate. But unlike agriculture, they produce almost no local economic circulation once construction ends.
Revenue that stays versus revenue that leaves
The critical distinction is not between old and new industries. It is between revenue that stays in the valley and revenue that leaves.
Revenue that stays
Agricultural operations distribute revenue across dozens of local participants: feed growers, equipment dealers, veterinarians, mechanics, truckers, processors, banks, and worker families. The dairy manufacturing sector carries an output multiplier of 2.93, according to the University of Idaho's IMPLAN-based analysis1, meaning nearly three dollars of total regional activity for every dollar of processing revenue. A dairy farmer's feed purchase becomes the feed grower's revenue, becomes the equipment dealer's sale, becomes the mechanic's paycheck, becomes the grocery store's receipt. Each transaction generates the next. This is revenue that stays in the valley, changing hands multiple times, building wealth across a broad base of families and firms.
Revenue that leaves
When farmland converts to a data center or utility-scale solar installation, the revenue structure inverts. Property tax and a handful of operational wages stay local. The remaining value flows to distant investors, corporate headquarters, equipment manufacturers, and tax equity structures through power purchase agreements, debt service, and corporate profit distribution. A data center may generate billions in cloud service revenue, but that revenue goes to the parent company. The local economy captures a fraction: property tax (often reduced by incentive agreements), a small number of specialized wages, and minimal local procurement. The community hosts the infrastructure; someone else's economy benefits from the output.
Why a small percentage matters: the coupling problem
To understand why 3 to 7 percent matters, you have to understand coupling. A coupled economy is one where the components depend on each other. The dairy depends on the feed grower. The feed grower depends on the irrigation infrastructure. The equipment dealer depends on both. The processor depends on the dairy. The worker's family depends on all of them. When these relationships are tight and local, the system is strong. But when one link weakens, the stress transfers to every other link.
The Magic Valley is one of the most tightly coupled agricultural economies in the United States. The University of Idaho's 2018 analysis documented total regional output of approximately $20 billion, a gross regional product of $7.4 billion, and just under 102,000 jobs1. Farm sales reached $3.8 billion, nearly 50 percent of Idaho's total farm cash receipts. Forty-five percent of cultivated acres in the region grow alfalfa hay and silage corn, the primary feeds for the dairy operations that anchor the economy.
When the system is this tightly coupled, removing even a small amount of productive land does not create a small problem. It creates a systemic one. The dairy manufacturing multiplier of 2.931 means that every dollar lost in dairy processing cascades to nearly three dollars of lost regional activity.
How the cascade works
The cascade follows a specific sequence. It begins with the land and works outward through every coupled sector.
Stage 1: Feed supply contraction. A dairy consuming 40 to 50 tons of feed daily requires roughly 2,000 to 2,500 acres of feed production land. When productive feed acres convert, remaining dairies must source from greater distances. Transportation costs rise immediately. A hay bale costing $15 to deliver from 15 miles away may cost $25 or more from 80 miles away. For a dairy consuming 40 tons daily, this differential accumulates across every load, every day, every month.
Stage 2: Dairy margin compression. Dairy operations typically run on profit margins of 5 to 10 percent. A feed cost increase of 10 percent from distant sourcing can eliminate the entire margin. Operations that were profitable become unprofitable. The weakest operations exit first, but their exit reduces the customer base for local feed suppliers, equipment dealers, and veterinary services, setting the stage for the next failure.
Stage 3: Vendor threshold collapse. Agricultural vendors exist because enough farms in a geographic area create viable customer bases. Research published through the American Farmland Trust's Farmland Information Center examined critical mass thresholds for agricultural support services4. The finding: below certain acreage levels, support businesses become economically unviable. Feed mills close. Equipment dealers consolidate into distant regional centers. Veterinary clinics reduce staff. This is not gradual. It accelerates. When one vendor closes, remaining vendors lose customers, pushing them closer to their own thresholds.
Stage 4: Processor vulnerability. Glanbia processes hundreds of millions of pounds of milk annually from regional dairies. Chobani's Twin Falls facility processes approximately 900 million pounds of milk per year. Lamb Weston processes over 1 billion pounds of potatoes annually. These processors located in the Magic Valley specifically because of concentrated production. When supply declines, processors face higher per-unit costs, reduced throughput, and the economic logic of relocation. The University of Idaho analysis found that 12 percent of the county-wide alfalfa crop is already at risk in current development scenarios1.
Stage 5: Workforce displacement and community erosion. When dairies close, vendors shut down, and processors reduce operations, the jobs disappear. But the damage extends beyond paychecks. Worker families are the multiplier in human form. They rent apartments, buy groceries, enroll children in schools, attend churches, and pay property taxes. When these families leave, school enrollment drops, tax revenue declines, and the social infrastructure that holds communities together weakens. This is not an abstraction. Idaho Department of Labor data documents the relationship directly: agricultural employment drives population stability, which drives school enrollment, which drives public service capacity.
Impermanence syndrome: how perception accelerates loss
Research published in the journal Choices (American Agricultural Economics Association) documented a phenomenon called impermanence syndrome6. When farmers perceive that development pressure will eventually absorb their land, they stop investing. They defer equipment purchases, skip soil improvements, reduce infrastructure maintenance, and begin planning exit strategies. The research found that for every acre of prime farmland urbanized, up to another acre becomes idle due to this disinvestment effect6.
This means that the 3 to 7 percent of farmland actually converted is not the total impact. The perception of conversion pressure causes additional acreage to become functionally unproductive. When a county approves two or three large conversion projects, the signal to every remaining farmer is clear: this land is being consumed. The rational economic response is to disinvest, which accelerates the decline beyond the acres physically converted.
The fiscal reality: what the studies show
The American Farmland Trust has conducted Cost of Community Services studies across 263 communities in 25 states over multiple decades5. The findings are remarkably consistent. Working and open space lands, including farmland, cost local governments $0.37 in services for every dollar of revenue they generate5. Residential development costs $1.16 for every dollar of revenue5. Commercial and industrial lands cost $0.30 per dollar of revenue5.
Farmland is not a burden on local budgets. It is a net fiscal contributor. When productive farmland converts to uses that demand infrastructure, roads, water systems, and services, the fiscal equation can worsen. The conversion removes a net fiscal contributor and may replace it with a use that demands more from public systems than it returns in revenue, particularly if tax incentive agreements reduce the new project's property tax contribution.
Coupling and decoupling: the structural choice
The 3 to 7 percent theory is fundamentally about what happens when a coupled system begins to decouple.
A coupled economy is one where economic relationships are local, mutual, and reinforcing. The dairy buys feed from the local grower. The grower buys equipment from the local dealer. The dealer employs mechanics who spend wages at local businesses. The dairy ships milk to the local processor, which employs hundreds of workers whose families sustain schools, churches, and civic institutions. Revenue circulates. Wealth distributes. Families form. Culture develops. Communities stabilize.
A decoupled economy is one where economic relationships are distant, one-directional, and extractive. A solar facility sends electricity to a distant utility through a power purchase agreement. Revenue flows to an out-of-state project owner. Equipment was manufactured overseas. Construction workers arrived temporarily and left. Two or three permanent employees maintain the facility. There is no feed chain. There is no vendor ecosystem. There are no processing jobs. There are no worker families enrolling children in schools. The land produces energy, but it does not produce community.
When farmland converts, the community does not merely lose acres. It loses coupling. It loses the relationships, the transactions, the dependencies that make the multiplier work. And because the system is interdependent, losing a small piece can begin the process of decoupling the whole.
Industry that builds families and culture
Agriculture in the Magic Valley is not merely an industry. It is the industry that built the families and culture of these communities. Hispanic and Latino families arrived generations ago to work in agriculture and have become integral to the social fabric of the valley. Multigenerational farming families have roots spanning 50, 80, or 100 years. These families did not just earn wages. They formed communities. They built churches. They coached youth sports. They served on school boards. They created the civic infrastructure that makes these places more than coordinates on a map.
This is what coupling looks like at the human level. Stable agricultural employment produces stable families. Stable families produce stable communities. Stable communities produce the institutions, the culture, and the social capital that no amount of tax revenue from an extraction project can replace.
When a dairy closes because feed costs rose because farmland converted, the loss is not measured only in dollars. It is measured in the families that leave, the school enrollment that drops, the Little League team that cannot field a roster, the church that closes, the volunteer fire department that cannot recruit. These are the consequences of decoupling, and they begin with the land.
Applying the theory to Jerome County
Jerome County encompasses approximately 155,000 acres of cultivated farmland, with roughly 40 percent of the county's 387,000 total acres in agricultural production. At a 3 percent threshold, the loss of approximately 4,650 acres could begin triggering cascade effects. At 7 percent, the loss of approximately 10,850 acres could push the system toward structural failure.
These are not large numbers. A single large solar installation can consume 500 to 2,000 acres. A data center campus and its buffer zones can consume 400 to 600 acres. Two or three large projects in a single county could approach or exceed the 3 percent threshold. Five to seven projects could approach 7 percent.
The concern is not that any single project will destroy the agricultural economy. The concern is that incremental conversions, each appearing manageable in isolation, accumulate to cross thresholds that trigger systemic harm. Each conversion decision is made individually, but the consequences are collective and compounding.
| Farmland Lost | Jerome County Acres (approx.) | Likely Impact |
|---|---|---|
| 1-2% | 1,550-3,100 | Manageable. Feed sourcing adjusts. Costs rise marginally. Impermanence syndrome begins among nearby farmers. |
| 3-5% | 4,650-7,750 | Feed costs rise noticeably. Marginal dairy operations begin evaluating exit. Some vendors experience reduced demand. Impermanence syndrome spreads. Property values for remaining farmland become uncertain. |
| 5-7% | 7,750-10,850 | Feed supply chain stress becomes acute. Multiple dairy operations face margin crisis. First vendor closures occur. Processing facilities begin evaluating supply reliability. Worker families begin relocating. School enrollment measurably declines. |
| 7-10% | 10,850-15,500 | Cascade accelerates. Dairy exits trigger further vendor closures. Processor throughput declines. Tax base erosion visible. Infrastructure maintenance deferred. Community institutions weaken. |
| 10%+ | 15,500+ | Structural transformation. The agricultural economy begins transitioning to a fundamentally different model. Remaining operations face economic isolation. Recovery becomes impractical. |
What the theory does NOT claim
This theory does not claim that 3 to 7 percent is a universal threshold applicable to every agricultural region. The threshold depends on the degree of coupling, the diversity of the agricultural base, the proximity of alternative feed sources, the financial strength of existing operations, and the structure of the local vendor ecosystem. A loosely coupled region with diverse agriculture and multiple feed sources may tolerate higher losses. A tightly coupled region like the Magic Valley, heavily concentrated in dairy with interdependent feed production, may be more sensitive.
The theory also does not claim that all farmland conversion is harmful. Conversion of degraded, non-irrigated, or marginal land to other uses may have negligible impact on the agricultural system. The concern is specifically about conversion of productive irrigated farmland on Class I and Class II soils that is actively coupled to the dairy and feed production system.
Finally, the theory does not claim that the exact percentage is known with precision. The 3 to 7 percent range is derived from the structural analysis presented in this series, informed by research on critical mass thresholds, impermanence syndrome, and the specific economics of the Magic Valley. More rigorous empirical research, including formal input-output modeling specific to Jerome County and the broader Magic Valley, would refine this estimate. The theory is offered as a framework for honest policy discussion, not as a proven mathematical constant.
Why this matters for policy
County commissioners, planning boards, and citizens evaluating farmland conversion proposals need a framework for understanding cumulative impact. Individual project approvals are evaluated one at a time, but their effects accumulate. The 3 to 7 percent theory provides that framework.
If a county has already approved projects converting 2 percent of productive farmland, the next project is not being evaluated against a clean baseline. It is being evaluated against a system that has already absorbed stress. The question is not whether this one project will destroy the agricultural economy. The question is whether this one project, added to all previous conversions, pushes the system closer to or past a threshold from which recovery becomes impractical.
Policy that fails to account for cumulative impact is policy that approves each individual step toward a cliff while insisting that no single step was the problem.
Questions for elected officials
- What is the total acreage of productive irrigated farmland that has been converted to non-agricultural use in your county over the past 10 years, and what percentage of total productive farmland does this represent?
- Has your county established a threshold, in acres or percentage, beyond which further farmland conversion would be considered harmful to the agricultural economy? If not, on what basis are individual conversion decisions being evaluated?
- When a farmland conversion proposal is presented, does the economic analysis account for cumulative impact across all previous and pending conversions, or does it evaluate the project in isolation?
- What analysis has been conducted on the feed supply chain for the dairy operations in your county, and at what point would local feed sourcing become economically unviable?
- If 5 percent of productive farmland in your county converted to non-agricultural use over the next five years, what would the projected impact be on dairy operations, vendor businesses, processing throughput, agricultural employment, and school enrollment?
Questions for the public
- Do you know how much productive farmland in your county has been converted in the past decade? Has anyone tracked whether the local agricultural economy has changed as a result?
- When a solar project or data center is proposed on irrigated farmland, do you consider it as one isolated decision, or as part of a cumulative pattern that may be approaching a tipping point?
- How many local businesses in your community exist because farms exist? Equipment dealers, feed suppliers, veterinary clinics, trucking companies, mechanics? What would happen to those businesses if the agricultural base contracted by 5 to 10 percent?
- Do you distinguish between industries that build families and culture in your community and industries that extract value from the land without building community? Which model do you want your county's future built on?
- If current conversion trends continue, what will your community look like in 20 years? Will the families, the schools, the churches, and the businesses still be there, or will they have followed the farmland out?
References
1 University of Idaho Extension. "Contribution of Agribusiness to the Magic Valley Economy, 2018." Extension Bulletin BUL 1005. Documents Magic Valley total output (~$20 billion), GRP ($7.4 billion), employment (102,000 jobs), and agricultural share (42% of jobs, 59% of sales). Reports dairy manufacturing multiplier of 2.93 and identifies 12% of county-wide alfalfa crop at risk in development scenarios. ↩
2 American Farmland Trust. "Farms Under Threat: The State of the States." 2020. Projects Idaho will lose 113,075 acres by 2040, with 83% on the state's best agricultural land. Documents conversion rate 96 times faster than previously projected. ↩
3 Idaho Dairymen's Association. Industry statistics, 2024-2025. Reports 350 dairy operations producing 18.26 billion pounds of milk (2025), $3.9 billion in farm-gate receipts, direct economic impact of ~$7 billion, total impact exceeding $11 billion, supporting 33,000 jobs and $155 million in state and local taxes. ↩
4 American Farmland Trust, Farmland Information Center. "Critical Mass: Does the Number of Productive Farmland Acres Matter?" Farmland Information Center Technical Report. Research on whether minimum acreage thresholds exist for sustaining agricultural support service networks, modeling critical mass levels for vendor viability. ↩
5 American Farmland Trust, Farmland Information Center. "Cost of Community Services Studies." Compilation of 263 studies across 25 states. Consistently finds agricultural and open space lands cost $0.37 in services per dollar of revenue; residential development costs $1.16 per dollar of revenue; commercial and industrial lands cost $0.30 per dollar of revenue. Idaho Farm Bureau Federation reports Census of Agriculture data (2017-2022) showing decline from 11.69 million to 11.55 million acres (144,000 acres lost), loss of approximately 14 square miles per year. ↩
6 American Agricultural Economics Association. "Desirability, Challenges, and Methods of Protecting Farmland." Choices Magazine. Documents impermanence syndrome: when farmers perceive development pressure, they disinvest; for every acre of prime farmland urbanized, up to another acre becomes idle due to farmer disinvestment. ↩
10 Deseret News. "How data centers are reshaping Idaho's farm country." March 2026. Documents Meta's $800M data center (485 acres near Kuna), Gemstone Technology Park ($1B, 620 acres). Capital Press, Idaho Statesman, Magic Valley Times-News. Various reporting, 2024-2026. Reports on Jerome County solar proposals (Cat Creek Energy, 554 acres south of Eden; 405-acre/150,000-panel project), county energy ordinance debates, and regional data center developments. ↩
11 Gemstone Technology Park, Kuna ID: 620 acres farmland rezoned for data center (Diode Ventures/Black & Veatch), estimated 600-800 MW. ↩