On Tuesday, Americans deliver their verdict on Donald Trump’s second term, and on the House the models speak with one voice: the average of 31 forecasts gives Democrats some 239 seats, with probabilities ranging from 80% to near-certainty. The Senate is genuinely contested; the consensus points to 52 Democratic seats and roughly a two-in-three chance of control, but with a spread from 49 to 56 and knife-edge races in Alaska, Kansas and Texas. Thus a Democratic House is the base case; a clean sweep is probable, but not assured.

How would Trump react to defeat? The domestic playbook has already been previewed. The Justice Department has sued pre-emptively to block certification of Alaska’s Senate race should the Republican incumbent lose narrowly — reportedly the first federal attempt to halt certification since the Civil Rights era — while Trump conditions his acceptance of the results on their being “honest”. His allies are openly preparing county-by-county certification fights. Yet certification is a state prerogative, and the courts have already pushed back; as in 2020, litigation is more likely to delay and delegitimise than to overturn. The damage, in other words, would be institutional erosion rather than reversal.

The external channel is more dangerous. According to Axios, a US official warns that an “embarrassed or disenfranchised” Trump would do something bold — namely, a large-scale campaign against Iran’s energy, infrastructure and nuclear targets; Gulf allies have reportedly been told to brace for a short, intense round as early as 4 November. A caveat is warranted: the White House denies any link, and the strike plans pre-date the vote. In fact, the election may determine the timing and intensity of the offensive rather than whether it happens — with Rubio and Hegseth pressing for it, and Vance, Witkoff and Kushner for talks.

Hence the frantic diplomacy. Qatar, alongside Pakistan, has tabled a seven-day draft built on daily reciprocal confidence-building steps and multiple guarantors; its mediators spent more than twelve hours in Tehran on Friday, in constant contact with the White House. The obstacle is sequencing: Washington wants nuclear concessions, starting with the enriched-uranium stockpile, at the outset; Tehran wants the naval blockade lifted and sanctions eased first. Since June’s interim agreement quickly unravelled, scepticism is warranted; a deal lands before Wednesday or, quite possibly, not at all.

Meanwhile, Israel’s 27 October election has produced deadlock: Netanyahu’s bloc holds 53 seats, Eisenkot’s 48, and the Arab parties a record 19. Arithmetically, Eisenkot’s only path to 61 runs through both Ra’am and the Joint List — a 67-seat majority, as Ra’am’s eight alone fall short. Yet three of the four parties in his bloc pledged never to govern with Arab parties, and Liberman’s refusal either to sit with them or to rely on their support may by itself sink the numbers. The 2021 Bennett–Lapid experiment proves the formula possible, and fragile. President Herzog, for his part, favours a unity government, which Eisenkot rejects if Netanyahu remains Prime Minister.

Therefore the likeliest outcome is a prolonged caretaker Netanyahu — precisely as Washington decides whether to resume the war. A wounded American president and an unseated, yet still governing, Israeli prime minister may find in Iran a common distraction; the mediators have days, not weeks, to deny them that option.