Serious Courses! In 2025, America’s average tariff rate climbed to roughly 16%—the highest since 1936, according to the Yale Budget Lab. The policy came with four promises: shrink the trade deficit, bring factories home, raise hundreds of billions in revenue, and make foreign exporters, not American shoppers, pay the bill.

Then, on February 20, 2026, the Supreme Court ruled 6–3 in Learning Resources v. Trump that the law behind most of those tariffs—the International Emergency Economic Powers Act (IEEPA)—does not give the President the power to impose them. The so-called reciprocal tariffs and the fentanyl tariffs were struck down. A temporary 15% replacement surcharge expired on July 24, and the courts have ordered roughly $165 billion in collected duties refunded to importers.

Not everything went away. Tariffs built on other laws—Section 232 (national security: steel, aluminum, autos and more) and Section 301 (unfair trade practices, mainly China)—are still in force, and Yale puts the average rate at about 11% today. That makes this a natural moment to open the ledger: we now have well over a year of official data on what the tariffs did, and a court ruling that partly unwound them.

Every figure below comes from a named source: the Bureau of Economic Analysis (BEA), the Census Bureau, the U.S. Treasury, the Bureau of Labor Statistics (BLS), the USDA, or published research.

1. The Big Picture: Why a Border Tax Can’t Fix a Savings Gap

Economists of nearly every stripe start from one accounting identity. A country’s trade balance equals what it saves minus what it invests:

“National saving” counts households, businesses, and government together. When a country spends and invests more than it saves—as the U.S. does, with large federal deficits and modest household saving—the difference has to be borrowed from abroad. That borrowing shows up, by definition, as a trade deficit. A tariff makes particular imports more expensive, but on its own it doesn’t make Americans save more or the government borrow less. It changes what we import and where it comes from more than how much.

┌──────────────────────────────────────┐
│ Large federal deficits + modest      │
│ household saving                     │
└──────────────────────────────────────┘
                   ▼
┌──────────────────────────────────────┐
│ Spending exceeds saving: the gap     │
│ is borrowed from abroad              │
└──────────────────────────────────────┘
                   ▼
┌──────────────────────────────────────┐
│ Tariffs raise the price of specific  │
│ imports; buyers switch suppliers     │
│ (China → Vietnam, Mexico, India)     │
└──────────────────────────────────────┘
                   ▼
┌──────────────────────────────────────┐
│ Result: the overall deficit          │
│ persists; its country mix shifts     │
└──────────────────────────────────────┘

The data fits that picture. In August 2026 the U.S. ran a $105.6 billion trade deficit in goods and services—a $136.6 billion goods deficit, partly offset by a $31.0 billion services surplus—the widest monthly gap since March 2025 (BEA and Census, released October 6, 2026).

There’s a wrinkle worth being honest about: for 2026 so far, the deficit is 19.9% smaller than over the same months of 2025. But 2025 was badly distorted—importers rushed goods in ahead of each tariff deadline, inflating the early-2025 deficit to record levels. Comparing against that surge flatters 2026. What the data doesn’t show is a lasting structural turn toward balanced trade.

What about the dollar? Textbook theory says tariffs should push the dollar up, cutting into exports and offsetting part of the effect. In practice the opposite happened: the widely quoted ICE U.S. Dollar Index (DXY) fell about 10.8% in the first half of 2025, its worst first half since 1973, as markets reacted to policy uncertainty. The Federal Reserve’s broader, trade-weighted index, shown below, fell 7.6%. A cheaper dollar should help exporters—and the deficit still didn’t close.

2. The Revenue Surge—and the Refund Bill

On revenue, the tariffs delivered. Customs duties brought in $194.9 billion in fiscal year 2025, up from $77.0 billion the year before—a 153% jump (Monthly Treasury Statement). In fiscal 2026, which ended September 30, net customs receipts fell 11% to $173.2 billion.

That 2026 figure hides a big number underneath. Treasury collected $317.8 billion in gross duties over the year but paid back $144.6 billion in refunds, nearly all of it after the Supreme Court ruling. With about $165 billion in IEEPA refunds ordered in total—and a class of importers who never sued certified on October 8—some of the revenue the tariffs raised is on its way back out.

Who actually paid? Not foreign governments. A tariff is paid to U.S. Customs and Border Protection (CBP) by the importer of record—the American company bringing the goods in. What happens next depends on how much of that cost each business along the chain passes on.

Here’s how a 20% tariff moves through the chain if every business keeps its usual percentage markup. (U.S. duties are charged on the goods’ value at the foreign port—freight and insurance aren’t taxed.)

                        Before  20% tariff
Factory price          $100.00     $100.00
+ Freight, insurance    $15.00      $15.00
+ Duty (20% of $100)         —      $20.00
= Landed cost          $115.00     $135.00
+ Importer (25%)        $28.75      $33.75
= Wholesale            $143.75     $168.75
+ Retailer (40%)        $57.50      $67.50
= Shelf price          $201.25     $236.25

$20 duty → $35 higher shelf price (+17.4%)

That’s the worst case for shoppers. Real businesses rarely pass everything through right away—they squeeze suppliers, accept thinner margins, or wait. The best direct measurement comes from Alberto Cavallo and colleagues at the Harvard Business School Pricing Lab, who track millions of retail prices: they estimate that seven months in, about 25% of the tariffs had reached store shelves, with imported goods’ prices rising roughly twice as much as comparable domestic goods. The rest was absorbed, for now, by importers, retailers, and foreign suppliers.

3. On the Ground: Three Case Studies

Cars: A Supply Chain That Crosses the Border Again and Again

North American car-making is one integrated system. The Cato Institute estimates that a single component can cross the U.S., Canadian, and Mexican borders up to eight times before final assembly—and a tariff can apply at each crossing.

[Metal]
   │
[Cast: Canada]
   │ ◄── border crossing (duty)
[Machine: U.S.]
   │ ◄── border crossing (duty)
[Assemble: Mexico]
   │ ◄── border crossing (duty)
[Car: U.S.]

Those auto tariffs rest on Section 232, so the Supreme Court ruling didn’t touch them. The Anderson Economic Group estimates that tariffs on Canadian and Mexican auto trade alone cost about $1,600 for every vehicle assembled in the U.S. in 2025—before counting steel and aluminum duties or parts from Asia and Europe. Rebuilding a stamping or casting plant takes years and billions of dollars, so in the short run automakers mostly paid the duties and raised prices or trimmed margins rather than moving production.

Electric vehicles face an extra squeeze. Battery production depends on refined lithium, graphite, and rare-earth magnets processed overwhelmingly in China. Taxing those inputs raises the cost of building EVs and batteries in the U.S.—working against the goal of growing that industry here.

Phones and Laptops: The Exemption That Proves the Point

Smartphones and laptops depend on chip fabs in Taiwan, display makers in South Korea, and assembly clusters across Asia. The U.S. doesn’t make them at consumer scale, and can’t start quickly. Tellingly, smartphones and computers were exempted from the reciprocal tariffs in April 2025—an admission that a tariff can’t protect an industry that isn’t here.

What tariffs did change was the label on the box. Brands moved final assembly from China to India and Vietnam, while the high-value chips and screens inside still came from the same Asian suppliers. That kind of rerouting is exactly what the accounting identity predicts: the source of imports shifts, but the total doesn’t shrink.

The Dinner Table: Olive Oil and Soybeans

Some imports have no domestic substitute at all. According to the USDA, imports supply more than 98% of the olive oil Americans use; California’s groves, the country’s main source, can’t scale up on a tariff’s timeline—new trees take years to bear fruit. When European goods faced a 15% tariff under the 2025 U.S.–EU trade framework, there was nowhere else to buy from, and food tariffs weigh most heavily on lower-income households, which spend a bigger share of their budgets on groceries.

Trade also runs the other way. When the U.S. raised tariffs, China retaliated against American farm exports and stopped buying U.S. soybeans for much of 2025. To keep farms afloat, the USDA launched a $12 billion Farmer Bridge Assistance program—$30.88 per acre for soybeans, a rate the American Soybean Association said won’t cover the losses. In effect, tariff revenue collected from American importers was partly recycled to American farmers hurt by the retaliation.

4. What It Costs a Household

Adding it all up, the Yale Budget Lab estimates that the tariffs in force today cost the average household about $1,100 a year, and raise the overall price level by about 0.7%. That estimate is lower than earlier ones for two reasons: the Supreme Court ruling removed many tariffs, and Yale updated its model in July 2026 to account for shoppers switching away from taxed goods.

Roughly half or more of what the U.S. imports isn’t finished consumer goods at all—it’s industrial supplies and equipment that American factories use to make things (St. Louis Fed, 1999–2019 data). So part of that household cost arrives indirectly, through higher prices for American-made products built from taxed parts.

5. The Scorecard

6. Should the Tariffs Stay?

The answer depends on what you think tariffs are for.

The case for them is strongest when the goal is strategic rather than economic. Robert Lighthizer, the U.S. Trade Representative in Trump’s first term, argues in No Trade Is Free (2023) that decades of open trade let state-subsidized foreign competitors hollow out American industry, and that tariffs are a necessary defense—especially for supply chains the country needs in a crisis, like defense manufacturing, chips, and critical minerals. On that view, higher prices are the cost of resilience.

The case against broad tariffs is strongest on the numbers above. Most economists, from former White House economic advisers to trade researchers at the Peterson Institute, make two points: the trade deficit is set by saving and investment, not by tariffs; and taxing imported parts and equipment handicaps the very American manufacturers the tariffs are meant to help.

A data-driven middle path looks something like this:

Used as a scalpel—targeted, temporary, tied to a clear strategic goal—tariffs can work. Used as a hammer across everyday commerce, the ledger so far shows American importers, manufacturers, farmers, and households paying most of the bill.

Going Deeper: More Videos

• The Political Economy of Tariffs — Marginal Revolution University: why tariffs persist even when they cost more than they save—concentrated winners, scattered losers.

• Why Trump’s Tariff Defeat Isn’t Over Yet — Bloomberg Television: the fight by more than 330,000 importers to get about $166 billion back, and the damage refunds can’t undo.

• Can I get my tariff money back now? — NPR’s The Indicator: unpacking the Supreme Court decision the day after it came down.

• Why the Supreme Court ruled against Trump’s tariffs — PBS NewsHour: the legal reasoning behind the 6–3 decision.

• The Little Known Supreme Court Doctrine That Helped Strike Down Trump’s Tariffs — Bloomberg Television: the “major questions” doctrine explained.

• Tariffs vs. Quotas — Marginal Revolution University: what tariffs and quotas do to prices and quantities, and the “deadweight loss” they create.

• The Smoot-Hawley Tariff — Marginal Revolution University: the last time U.S. tariffs were this high, and the retaliation that followed.