Freeenterpriseland | Thoughts & Commentary

A subsidy is a transfer first and a market intervention second. The transfer is the easy part to see: the check, the credit, the guarantee, the price floor. The intervention is everything that happens after. Relative prices move. Entry costs move. The set of firms able to compete holds steady or shrinks. Incumbents live in the gap between those two things.

This article reads that gap in primary documents. The CHIPS Program Office rules. Treasury’s Section 45X final rule. USDA’s crop insurance and sugar program texts. The WTO record on export subsidies. The pattern holds across all of them. Programs written for stated public purposes rely on certain mechanics — scale thresholds, definitional line-drawing, uncapped premiums, assigned quotas — and those mechanics tilt toward firms already in the market. Sometimes the tilt is a deliberate design choice. Sometimes it is an accident of drafting. The documents let you tell the difference, if you read them for who they admit and who they price out.

What Distortion Means in a Subsidized Market

Answer first: a subsidy distorts a market when it changes what firms do, not just what they earn. Four channels recur across programs. Learn to spot them and most subsidy fights become legible.

The transfer. Money moves from the public ledger to private hands. This is the visible part, the part that makes headlines. It is usually the least interesting part.

The threshold. Eligibility conditions — minimum investment, matched funds, application burdens — are fixed costs of participation. Fixed costs filter out small firms before anyone evaluates their ideas.

The definition. Rules draw lines. What counts as production, as domestic, as a component. Lines allocate benefits. Whoever sits on the winning side of a line keeps the credit.

The capitalization. When a benefit attaches to a fixed asset — land, quota, a fabrication plant — the asset’s price rises to meet it. Early holders capture the transfer. Late entrants pay it back as higher prices or rents.

Each channel leaves a signature in the documents. The rest of this article reads five of them.

Scale Thresholds: The CHIPS Act’s Built-In Tilt

The CHIPS and Science Act of 2022 (P.L. 117-167) appropriated $52.7 billion for the semiconductor sector. About $39 billion went to manufacturing incentives. Section 48D added a 25 percent investment tax credit for semiconductor facilities and equipment. The statute stated the purpose plainly: national security, supply chain resilience, regional development. The Commerce Department’s CHIPS Program Office wrote the operating rules, now codified at 15 CFR part 231.

Now look at the finalized awards from late 2024. Intel: roughly $7.9 billion. TSMC: $6.6 billion. Samsung: about $4.7 billion, down from the $6.4 billion preliminary figure. Micron: about $6.2 billion. GlobalFoundries: $1.5 billion. Every named recipient already builds leading-edge or near-leading-edge chips. Every one already had fabs in the United States or had broken ground. The money followed the incumbency.

The application as a filter

Ask why, and the mechanics answer.

First, the application itself. CHIPS funding required engineering and cost documentation, financial due diligence, workforce plans, and — for facilities seeking more than $150 million — an affordable childcare plan for workers. None of that is objectionable on its face. All of it is a fixed cost. A large firm staffs the application with lawyers, accountants, and program managers. A small firm pays the same fixed cost against a small award, or does not apply at all.

Second, matched capital. Recipients had to put their own money at risk alongside federal funds. That requirement reads as discipline, and in part it is. It also means only firms with access to billions can play. A fab costs more than most companies’ total market value.

Third, clawbacks. The guardrails rule the CHIPS Program Office issued in September 2023 barred recipients from expanding advanced capacity in countries of concern for ten years, on pain of returning the full award. A firm betting its whole balance sheet on one project cannot price that risk. A diversified incumbent can absorb it. Regulatory risk, like application cost, is a fixed burden — and fixed burdens weigh on small balance sheets more.

Fourth, additionality. Commerce said it would fund projects that would not happen but for the subsidy. That test selects for large, risky projects. Which is to say, for incumbents. The purpose and the filter point in the same direction.

The terms also move. In August 2025, Commerce converted the remaining portion of Intel’s award into an equity stake of roughly ten percent in the company. Set aside what you think of the transaction. The document lesson stands on its own: the instrument is administrative, and the government can change it after the fact. The firms that can absorb that kind of renegotiation are, again, not startups.

Definitions as Allocation: Treasury’s 45X Line Between Extraction and Production

The Inflation Reduction Act created a production credit in section 45X — a per-unit or percentage credit for manufacturing eligible clean-energy components in the United States. Treasury wrote the definitions. The definitions decided who got paid.

The December 2023 proposed rule drew the credit’s starting line after extraction. Digging critical minerals out of the ground was not production. Refining and processing were. The April 2024 final rule moved the line. Extraction costs became creditable for critical minerals.

Why does the line matter? Because allocation follows definition. Under the proposed line, the credit rewarded firms that refine. Under the final line, it also rewarded firms that mine domestically. The reversal was worth something specific to companies already holding U.S. extraction assets — firms already in the business, in other words. A firm planning to enter refining from scratch gained nothing from the reversal. A firm with an existing mine gained a new claim on existing operations.

This is line-drawing as subsidy allocation. It is the most common form of distortion in tax-credit programs, and it stays invisible unless you compare the proposed rule against the final.

The clean-vehicle credit in section 30D works the same way, one layer deeper. Treasury’s sourcing rules tied the credit to critical minerals extracted or processed in the United States or a free-trade-partner country, and to battery components assembled in North America. A December 2024 proposed rule added “foreign entity of concern” restrictions on top. Each test rewards a specific supply-chain shape. Firms with compliant supply chains already in place qualify. Firms that would have to build them face a choice: enter under the incentive, or wait.

Then the ground moved. The July 2025 reconciliation law ended the clean-vehicle credit for acquisitions after September 30, 2025, and put staggered termination dates on several other credits while leaving parts of the manufacturing credit standing. Repeal did not remove the distortion. It re-sorted it. Grandfathered projects keep their benefits. The next entrant faces a different rulebook. Whoever was inside when the music stopped holds the claim.

Capitalization: Why Farm Subsidies Raise the Cost of Farming

The Federal Crop Insurance Program is the largest single channel of farm support. The government pays, on average, about sixty percent of every premium dollar. The policies are sold and serviced by private companies — roughly fifteen approved insurance providers, with a handful holding most of the business. The Standard Reinsurance Agreement between USDA and those companies assigns most underwriting losses to the government while letting the companies retain underwriting gains. The taxpayer carries the downside. The incumbents on the delivery side keep a capped, protected version of the upside.

Note the payment-limit asymmetry. Commodity program payments carry an adjusted-gross-income ceiling of $900,000. Premium subsidies carry no comparable hard cap, and a waiver process lets high-income producers back into eligibility. The result is arithmetic, not argument. The largest operations insure the most acres at the highest coverage levels, so they collect the largest absolute subsidies, year after year.

Now the distortion. USDA’s Economic Research Service has published on this for years: program benefits capitalize into land values and cash rents. A subsidy raises the expected return to farming a given acre. Bidders and renters price that expectation in. Land prices rise. Rents rise. The person who owned the ground when the subsidy arrived captures the transfer as asset value. The person trying to enter farming today pays the subsidy back as a higher rent check.

That is capitalization, and it is the quietest large distortion in the farm portfolio. The check goes to the operator. The benefit goes to the landowner. The cost goes to the next operator. Nothing in the payment data shows this. Only the asset-price data does.

The 2019 trade payments as a planting signal

The Market Facilitation Program showed the other half of the mechanism: subsidies instruct behavior. USDA paid trade-war compensation at county-level rates tied to which covered crops a county planted. GAO’s May 2020 review (GAO-20-371) found the rates bore an uneven relationship to measured damage, and observed that the design gave farmers a reason to plant covered crops regardless of market demand. The rule told farmers what to plant. Farmers read the rule. That is what a price signal is — and the government had written one into the crop mix.

Quota Rents: The Sugar Program Assigns Incumbency

The U.S. sugar program is the rare subsidy that states its incumbency in operating tables. It has three parts. USDA lends to processors at a set rate per pound, and the loans are nonrecourse: if the price floor holds, the processor repays; if it does not, the processor forfeits sugar and keeps the money. Import tariff-rate quotas cap how much foreign sugar can enter at low duties. And marketing allotments divide the permitted domestic supply among existing processors — by name, in USDA’s published tables.

The combination keeps domestic sugar prices well above world prices, often near double. That spread is a quota rent. It accrues to whoever holds the allotment and the import share.

Ask the entry question. A new beet processor cannot simply build a plant and sell sugar. It needs allotment. USDA reallocates allotment when supply runs short, not when a competitor wants in. In tight years, USDA has instead raised import quotas to relieve pressure on food manufacturers, as it did in 2023 and 2024. The floor survives every adjustment. Entry never improves.

The program even has a pressure valve that confirms the design. The Feedstock Flexibility Program has USDA buy surplus sugar and resell it to ethanol plants as feedstock, at a loss, rather than let the price fall. The buyer of last resort is the government. The sellers are allotment holders.

Read the sugar program next to any tariff file and you can see why this site covers subsidies and trade actions together. A price floor plus a tariff-rate quota is a subsidy and a trade barrier performing one job: protecting the assigned position of the processors listed in the tables.

Subsidies Migrate: FSC, ETI, Section 199, FDII

Subsidies do not die when they lose in court. They migrate. The record here is unusually clean because the referee was international.