The tariff that boomerangs: Who bears the cost if the US targets buyers of Russian oil

The US has found a new way to pressure Russia: punish the countries that buy its energy. My modelling suggests that the biggest loser could be the US itself. The US-Russia diesel deal struck last Friday, three weeks after the law passed, shows why.

Sanctioning law

On 18 September 2026, the Lindsey O. Graham Sanctioning Russia and Iran Act of 2026 became law. On Iran, it simply extends the Iran Sanctions Act of 1996 until 2031. The real action is on Russia.

The Act requires tariffs of up to 500 per cent on US goods imports from Russia1. Its more distinctive feature is the use of trade policy to target Russia’s customers, turning tariffs into a tool of secondary sanctions. The law provides for tariffs of up to 100 per cent on goods imported into the US from:

• the five largest buyers of Russian crude oil or natural gas, by volume over the 12 months preceding enactment, if they knowingly make new purchases on or after 18 October 2026,
• the five leading countries facilitating Russian oil sanctions evasion.

These duties are additional to existing tariffs. Although the measures are mandatory on paper, implementation remains uncertain. The rate can be adjusted as countries increase or reduce their purchases of Russian energy. Countries whose Russian gas imports are below 15 per cent of Russia’s total gas exports may be exempt if they take significant steps to reduce them. The President can also waive duties in the national interest, subject to congressional reporting requirements.

The diesel deal: sanctions undercut from within

If there is one certain thing about President Trump’s trade policymaking, it’s that it is inconsistent. On the one hand, he announces tariffs on countries trading with Russia, and less than three weeks later, he announces a deal to release the sale of Russian diesel. This blog focuses on the consequences of the former, though noting the inconsistencies and that the diesel deal may affect the implementation of the tariffs.

In this blog, using my quantitative trade model,2 I ask a simple question: what would happen if the US applied the maximum rate, a 100 per cent tariff on all manufactured goods imported from China, India and Turkey?3 These three countries are the largest buyers of Russian crude oil, together accounting for around 92 per cent of Russia’s crude exports since December 20224, and their differing trade relationships with the US make the comparison informative. I then halve the rate to 50 per cent to test how sensitive the results are to the choice of rate.

The results are illustrative, not precise forecasts. They show the potential direction and magnitude of welfare changes for a three-country scenario.

A near-prohibitive tariff that the US pays for

The first result is the most striking: the tariff hurts the country imposing it. Under a 100 per cent tariff on all three countries, US welfare falls by 0.84 per cent5. This loss is net of the tariff revenue collected: the revenue does not come close to making up for the higher cost of imports.

This runs against a well-known argument in trade theory, that a large economy can gain from a moderate tariff by pushing down the prices its suppliers receive. That terms-of-trade gain exists here too, but at 100 per cent it is overwhelmed, as the duty is close to prohibitive. US imports from the three targets collapse by around 84 to 86 per cent. When trade is destroyed rather than taxed, there is little left to raise revenue from, while American households and firms lose access to low-cost goods and must re-source from more expensive suppliers. A moderate tariff imposed by a large economy might extract a transfer from foreign suppliers; a near-prohibitive one is likely simply to burn the gains from trade.

Table 1. 100% standalone tariff on China, India and Turkey6 

Note: Percentage change in welfare (real wages plus tariff revenue) and in real US imports from each partner. No retaliation, pre-tariff baseline.

Targets lose in proportion to their exposure, not size

All three countries in this scenario see their US exports collapse by similar amounts, but the burden does not follow economic size. China, the largest US supplier by far, loses the least: 0.24 per cent of welfare. This is likely because its economy is large and diversified enough to redirect lost sales to other markets. India (−0.58 per cent) and Turkey (−0.54 per cent) are hit more than twice as hard, likely because they depend more on the US market and have fewer alternative outlets to absorb the shock.

Trade shifts to third countries: the winners are elsewhere

If the US and the targets both lose, where do the gains go? The answer is to third countries that step into the gap. Mexico’s welfare rises by 0.72 per cent in the short run, Vietnam’s by 0.44 per cent and Bangladesh’s by 0.42 per cent, with US imports from Vietnam and Bangladesh rising by around a fifth. The lost imports from the targets are largely replaced by imports from other suppliers, redirecting global trade around the US at the US’s own expense.

The UK is among the beneficiaries, though more modestly. UK welfare rises by 0.11 per cent, as US imports from the UK increase by around 13 per cent. British exporters pick up some of the US market share vacated by China, India and Turkey.

Robustness: what if the rate were 50%?

As the Act sets a ceiling rather than a rate, I rerun the same scenario with a 50 per cent tariff (Table 2). The story is unchanged: the US still loses, the targets still lose, the ranking among targets is the same, and the same third countries, the UK included, still gain.

However, the effects of the tariff rate are nonlinear: halving the rate does not simply halve the damage. For the US, it cuts the short-run welfare loss by two-thirds, from 0.84 to 0.28 per cent. Two things likely drive this. First, a lower rate destroys less trade, so more revenue is collected and it offsets more of the rise in import prices. Second, each extra point of tariff likely does more harm than the one before. A modest tariff only pushes out imports that were barely cheaper than the alternatives, so little is lost. A very high tariff can push out goods that were much cheaper than anything else available, forcing American households and firms onto far more expensive substitutes. The US loss therefore falls steeply as the rate is lowered, which suggests the costs to the US are highly sensitive to the rate chosen.

Table 2. 50% standalone tariff on China, India and Turkey

Note: Percentage change in welfare (real wages plus tariff revenue) and in real US imports from each partner. No retaliation, pre-tariff baseline.

Put together, the comparison suggests that the choice between 50 and 100 per cent is mainly a choice about how much the US hurts itself. The pressure on the targets is largely in place at the lower rate.

The key takeaway

At these rates, a secondary tariff is likely to be an act of economic self-harm for the US that mainly benefits exporters in non-targeted countries. Even after counting the revenue it raises, the US loses more than China, the buyer whose purchases arguably matter most to Russia’s revenues. A large, diversified China absorbs the shock by redirecting its exports. The heaviest costs fall on the US, India and Turkey rather than China, as they are less capable of redirecting their exports. The displaced trade is spread across many non-targeted countries in the model, which pick up the US market share the targets lose. Mexico, Vietnam, Bangladesh and, more modestly, the UK are among those that gain, but they are illustrative examples rather than a full list of winners.. Halving the rate to 50 per cent sharply reduces the damage the US does to itself, but leaves most of the pressure on the targets in place.