India's Russian oil arrivals fell 19.8%. The tariff's role cannot be isolated

The immediate policy question comes from Washington. Supporters unveiled a revised measure on Tuesday, July 14, that would reportedly allow tariffs of up to 100% on the world's five largest purchasers of Russian oil or natural gas, with China and India at the top of the list identified by sponsors, according to the [Associated Press](https://apnews.com/article/graham-senate-ukraine-russia-sanctions-e0e22a2c90391ad527547093e07e3661). The official [Senate Foreign Relations Committee statement](https://www.foreign.senate.gov/press/dem/release/shaheen-blumenthal-graham-wicker-announce-agreement-on-legislation-to-hold-purchasers-of-russian-oil-accountable) issued four days earlier confirmed an agreement with the administration to advance updated legislation but did not publish those detailed terms.

India offers a recent comparison, but not a controlled experiment. [Executive Order 14329](https://www.whitehouse.gov/presidential-actions/2025/08/addressing-threats-to-the-united-states-by-the-government-of-the-russian-federation/) imposed an additional 25% duty on covered Indian goods because India was importing Russian oil; it took effect on August 27, 2025. A later [White House order](https://www.whitehouse.gov/presidential-actions/2026/02/modifying-duties-to-address-threats-to-the-united-states-by-the-government-of-the-russian-federation-04b2/) removed the duty for goods entered from February 7, 2026.

Across the five full calendar months between those dates, September through January, Russian seaborne crude arrivals averaged 1.466 million barrels a day. The immediately preceding five months averaged 1.829 million. That is a decline of 0.363 million barrels a day, or 19.8%, calculated from the [KSE Institute's Kpler-based monthly estimates](https://institute.kse.ua/wp-content/uploads/2026/04/russian_oil_tracker_eng_march_2026.pdf) and weighted by the days in each month using the [EIA definition of barrels per day](https://www.eia.gov/tools/faqs/faq.php?id=62&t=7).

A single before-and-after average hides the sequence. Arrivals were 1.578 million barrels a day in September and 1.608 million in October, then rose to 1.828 million in November before falling to 1.241 million in December and 1.091 million in January, according to the [KSE-Kpler series](https://institute.kse.ua/wp-content/uploads/2026/04/russian_oil_tracker_eng_march_2026.pdf). February, mostly after the tariff had ended, was lower still at 1.042 million.

The first two full tariff months therefore averaged 1.593 million barrels a day, below the pre-tariff rate but followed by a November rebound. The post-wind-down months of December and January averaged 1.166 million, the largest phase break in this short series. That timing is consistent with an effect from the later supplier sanctions, but consistency is not proof of cause.

The analogy is a road with several tollbooths installed close together. Traffic fell after the first toll appeared, but the sharpest fall came after another toll opened. Counting fewer cars at the end tells us the whole route became harder to use; it does not tell us how many cars each toll diverted.

First came the tariff. Its direct legal object was Indian merchandise entering the United States, not crude entering India: [the order](https://www.whitehouse.gov/presidential-actions/2025/08/addressing-threats-to-the-united-states-by-the-government-of-the-russian-federation/) added a 25% customs duty to covered Indian goods, with stated exclusions. The order did not levy that customs charge on India's Russian oil cargoes, and the available data do not identify the behavioral path from the tariff to refinery purchases.

Second came blocking sanctionsblocking sanctionsBlocking sanctions freeze a targeted person's property within U.S. jurisdiction or U.S. control and generally prohibit dealings in it unless authorized, as [OFAC explains](https://ofac.treasury.gov/faqs/9). on suppliers. On October 22, the U.S. Treasury [designated Rosneft and Lukoil](https://home.treasury.gov/news/press-releases/sb0290), and [General LicenseGeneral LicenseA general license is a public OFAC authorization for specified transactions that sanctions would otherwise prohibit, provided its conditions are met, according to [OFAC](https://ofac.treasury.gov/faqs/7). 126](https://ofacofacOFAC is the U.S. Treasury office that administers and enforces U.S. economic and trade sanctions, according to its [official overview](https://ofac.treasury.gov/faqs/all-faqs)..treasury.gov/media/934706/download?inline=) authorized specified wind-down transactions only through November 21. Those sanctions changed which transactions involving the blocked companies U.S. persons could conduct without authorization.

Third came the EU's rule for refined fuel. From January 21, 2026, [EU Regulation 2025/1494](https://eur-lex.europa.eu/eli/reg/2025/1494/oj/eng) prohibited covered petroleum products made in a third country from Russian crude, subject to the regulation's rules and exceptions. Reliance said it stopped routing Russian crude into its export-oriented refinery ahead of that deadline, while Russian cargoes could still go to its domestic-market refinery, according to the company's statement reported by [Business Standard](https://www.business-standard.com/companies/news/reliance-ril-non-russian-crude-oil-export-refinery-us-sanctions-125112001339_1.html). That example shows why a national total can conceal different decisions inside the same refining group.

The KSE chart uses Kpler vessel-tracking estimates. [Kpler](https://www.kpler.com/product/commodities/cargo-analytics) describes a system built from vessel positions, cargo details, load and discharge locations, destination forecasts and port arrivals and departures. Those data are useful for measuring physical flows, but an arrival in one month can reflect a cargo loaded under an earlier commercial and sanctions environment.

The U.S. licenses issued in 2026 make that timing problem visible. [General License 133](https://ofac.treasury.gov/media/935101/download?inline=) authorized qualifying Russian oil already loaded by March 5 to be delivered to India through April 4. [General License 134C](https://ofac.treasury.gov/media/935641/download?inline=) later authorized qualifying Russian oil loaded by April 17 to be delivered through June 17. A delivery series can therefore move after the legal rule or purchase decision that set the cargo in motion.

This does not make vessel data unreliable. It means the date on the chart is the date of physical arrival, not a timestamp for every earlier contracting, financing, loading or compliance decision. A causal study would need cargo-level dates and counterparties, not only monthly national totals.

A July 2 report based on Kpler data estimated India's June Russian crude arrivals at 2.61 million barrels a day, up from 1.87 million in May, according to the [Financial Express](https://www.financialexpress.com/policy/economy/russian-crude-imports-surge-39-to-2-61-million-bpd-make-up-52-of-indias-june-oil-intake/4281852/). We plot the June value separately because it comes from a later media report, while March through May are absent from the KSE monthly series used for the earlier comparison.

The rebound occurred after a severe Middle East supply disruption. The [International Energy Agency's March 2026 Oil Market Report](https://www.iea.org/reports/oil-market-report-march-2026) described the disruption as the largest in the history of the global oil market and projected a sharp March supply loss, while the OFAC licenses changed which already-loaded Russian cargoes could be delivered. Chronology alone cannot tell us whether June reflected the supply shock, the licenses, commercial discounts, refinery maintenance ending, normal volatility or some combination.

The June figure therefore rebuts only the strongest descriptive claim that the late-2025 decline continued without reversal. It does not show that the tariff failed, that sanctions failed or that either policy succeeded, because those judgments require an estimate of what arrivals would have been without each intervention.

The missing object is the counterfactualcounterfactualA counterfactual is an estimate of what would have happened without the policy, which the [World Bank's impact-evaluation handbook](https://www.worldbank.org/en/programs/sief-trust-fund/publication/impact-evaluation-in-practice) identifies as the missing comparison needed to measure an effect.: a credible estimate of India's Russian oil arrivals without the tariff but with everything else that actually changed. The [World Bank's impact-evaluation handbook](https://openknowledge.worldbank.org/bitstream/handle/10986/25030/9781464807794.pdf) explains why the observed before-and-after difference cannot supply that comparison by itself.

A stronger study would combine cargo-level arrivals and load dates with refinery-level purchases, supplier identities, prices, maintenance schedules, total Indian crude demand and comparable buyers not exposed to the India-only tariff. It would then test whether India's path changed differently after August 27 while allowing separate breaks for October 22, November 21 and January 21. Without those data and a defensible comparison group, a numerical tariff effect would be false precision.

The same caution applies to economic effects. This article measures physical oil arrivals, not U.S. import prices, Indian export orders, refinery margins, consumer fuel prices or Russian revenue. A separate [USITC study of 2018-2021 U.S. tariffs](https://www.usitc.gov/press_room/news_release/2023/er0315_63679.htm) found that U.S. importers bore nearly the full cost of the tariffs it studied, but those estimates concern different measures and cannot be transferred to the India tariff. They illustrate why who pays customs and who bears the wider economic burden are different questions.

The earlier episode shows that India's Russian crude arrivals were lower during the five full calendar months of the tariff and that the deepest decline overlapped later supplier sanctions. It does not supply a reliable percentage response that lawmakers can multiply by a new tariff rate, because the tariff's independent contribution was never identified.

The legal structure also changed. The 2025 measure was a presidential order aimed at covered Indian goods under IEEPAIEEPAIEEPA is the International Emergency Economic Powers Act invoked for the 2025 tariff, although the [Supreme Court held on February 20, 2026](https://www.supremecourt.gov/opinions/25pdf/24-1287_4gcj.pdf) that the law does not authorize a president to impose tariffs., and the [Supreme Court held on February 20, 2026](https://www.supremecourt.gov/opinions/25pdf/24-1287_4gcj.pdf) that IEEPA does not authorize presidential tariffs. The July proposal described by the [Associated Press](https://apnews.com/article/graham-senate-ukraine-russia-sanctions-e0e22a2c90391ad527547093e07e3661) is legislation, covers the five largest buyers identified under its terms and would allow rates up to 100%, with reported exceptions and presidential flexibility.

The defensible conclusion is narrow but useful: the 2025 India tariff coincided with a 19.8% decline in Russian crude arrivals across the chosen before-and-after windows. The data do not reveal the tariff's share of that decline, so using 19.8% as the predicted effect of the new proposal would confuse description with causation.