A Number That Did Not Change
Whichever party controls the next Senate, one number was never on the ballot. No party will hold 67 seats, the two-thirds majority the Constitution requires to ratify a treaty. That matters because, within days of the vote, US envoys are due back at the table with Russia and Ukraine, most likely in Abu Dhabi in mid-November, with an American security guarantee at the centre of the bargain. On 22 October Volodymyr Zelensky said any US commitment must be approved by Congress: Kyiv wants “not a piece of paper,” he said, but a guarantee that cannot be shrugged off like the 1994 Budapest Memorandum.
The easy conclusion is that without 67 votes, Ukraine will get only an executive agreement that the next president can tear up. That conclusion is wrong about the arithmetic and right about the outcome. The votes probably exist. What they would buy is the problem.
Treaty, Agreement, or Something In Between
Stay ahead of the geopolitical week.
MD Briefing delivers expert analysis across five global fronts — the Indo-Pacific, energy, geoeconomics, European security, and the Middle East — every Monday morning. Free.
A treaty needs two-thirds of the Senate. An executive agreement needs no vote at all, and the president who signs it, or the next one, can walk away from it. In between sits a congressional-executive agreement, approved by simple majorities in both chambers. Zelensky has spoken of a “vote in Congress” without saying which.
Washington has signalled all three. In December 2025 a senior US official said the administration was ready to send the Senate a guarantee text “based on Article 5” of the NATO treaty. The Paris Declaration of January 2026 described a US assistance obligation that Congress would need to approve, but whose triggering would be a political decision. By 20 October, US officials were promising guarantees “similar to” Article 5, which Zelensky took care to say is not the same thing.
Why Ratification Does Not Make It Korea
Start with the arithmetic, because it is better for Kyiv than the headline suggests. Supermajorities for defence treaties are not a relic: the Senate approved the Korea Mutual Defense Treaty 81 to 6 in January 1954, and Finland and Sweden’s NATO accession 95 to 1 in 2022. On Russia itself, the Senate passed the Lindsey O. Graham Sanctioning Russia Act 86 to 11 in August. Only one Republican, Rand Paul, voted no, alongside ten Democrats. A promise to defend Ukraine is a far bigger ask than sanctions, and progressives and Republican restrainers could combine against it. But 34 votes to block a guarantee that the president himself is sponsoring are hard to find. Control of the chamber barely moves that count.
So suppose the treaty passes. Would it make Ukraine’s ceasefire line hold like Korea’s? The text of the Korean treaty suggests not. Its operative clause commits each party to “act to meet the common danger in accordance with its constitutional processes” — language as discretionary as NATO’s promise to take “such action as it deems necessary.” The Korean armistice line has held for 73 years because 28,500 American soldiers stand behind it, not because of that sentence. A North Korean attack on the South would kill Americans on the first day. As MD’s analysis of why a peace deal remained elusive noted earlier this year, the hard questions in Ukraine have always been about what happens on the ground. The American role on offer there is a sensor network to verify breaches, with European troops kept away from the contact line.
Ratification also does not solve the problem it is supposed to solve, permanence. Presidents have ended ratified treaties on their own authority: George W. Bush withdrew from the Anti-Ballistic Missile Treaty in 2002, and Donald Trump left the Intermediate-Range Nuclear Forces Treaty in 2019 and the Open Skies Treaty in 2020. Congress has closed that door for exactly one alliance. The 2024 defence authorisation act bars a president from leaving NATO without two-thirds of the Senate or an act of Congress. Even that lock is being tested: in April, Trump floated leaving NATO over allies’ refusal to join the Iran war. A Ukraine treaty without such a clause would be easier to leave than the alliance it imitates.
This is why the Karabakh comparison holds even if the Senate votes. Russia’s 2020 guarantee to the Armenians of Karabakh was signed and backed by troops on the ground, yet it collapsed in 2023 because nothing obliged Moscow to act when its priorities changed. Its peacekeepers had no mandate to fight, and no domestic institution could force them to. A Ukraine guarantee written in “deems necessary” language, terminable by the president and backed by sensors rather than soldiers, has the same weakness, only with better paperwork.
The strongest objection is that ratification is not just paperwork. A treaty voted on by 67 senators raises the political cost of walking away; Budapest was never ratified and failed, while NATO was ratified and has held. That is true, and it is why a vote matters. But it cuts the other way too. Analysts at RAND point out that three successive presidents have declined to send US troops to defend Ukraine. A grand guarantee that Washington is unlikely to honour may weaken confidence in every other American commitment, NATO’s included. A vote that hides that gap would be worse than no vote at all.
The midterms therefore matter less for whether Ukraine gets 67 votes than for what the next Senate attaches to them. A chamber that wants the guarantee to bind would add a NATO-style withdrawal lock, reporting requirements, and some American presence in Ukraine, even a monitoring one. A chamber content to ratify whatever comes from Abu Dhabi would not.
Three Ways the Promise Gets Written
Base case — better paperwork, same gap (roughly 50%). Congress approves a guarantee within a year of a ceasefire, by treaty or a simple-majority agreement, with “Article 5-like” wording, no withdrawal lock and no US forces in Ukraine. This gives Kyiv a stronger legal claim than Budapest, but it leaves Moscow free to test the line with incremental probes, as Azerbaijan did in Karabakh, betting that no US president will treat a blockaded town as a reason for war.
Downside — Budapest II (roughly 30%). No vote is ever held. Talks stall over territory, or the administration settles for a political declaration pending a ceasefire Russia never fully accepts. The guarantee exists only in a signing statement. For governments and investors pricing Ukraine’s reconstruction, this is the outcome that makes political-risk insurance close to unobtainable.
Upside — Korea-lite (roughly 20%). A bipartisan group in the new Senate makes ratification conditional on two things: a clause requiring two-thirds of the Senate to withdraw, copied from the NATO law, and a standing US presence in Ukraine, even a small one. Moscow objects but accepts a ceasefire because sanctions under the August law bite. This is the only version in which the guarantee would bind a future president.
Probabilities reflect the author’s analytical judgment based on currently available information and are not statistical forecasts.
Watch the Noun, Then the Clause
No party will hold 67 seats, but that was never the obstacle. The votes for a Ukraine guarantee are there to be had. What a two-thirds vote cannot buy on its own is a commitment that survives a change of president, and a reason for Moscow to believe Americans would defend it. Without both, the deal from Abu Dhabi will be Karabakh with a Senate seal.
Watch the noun first. If the administration calls the outcome a “treaty,” the Senate will vote; if it says “agreement” or “framework,” it may never vote at all. Then watch whether senators from both parties start demanding a ratification vote, and whether any of them puts a withdrawal lock in the text. The Senate’s signature is the easy part. The clause that stops the next president from removing it is what would make the guarantee hold.
