Every November The Economist publishes The World Ahead, its guide to the coming year. This year’s edition is confident. The year 2027 is “certain” to be the hottest on record. Marine Le Pen is the “odds-on favourite” in France. Donald Trump “will not be impeached”, though members of his cabinet will be.

Tom Standage’s ten trends

Calls like these are easy to nod along with and hard to score. Is “odds-on” 60% or 90%? So we took ten of the magazine’s themes and wrote questions that a public record will settle on a known date. Prolepsia forecast each one between 13 and 17 November 2026. The numbers below are frozen at those values.

Here is the short version.

72% that the US House impeaches a sitting cabinet secretary in 2027. 38% that it impeaches Trump.

88% that America strikes Iran again in 2027. 34% that the two countries announce a nuclear deal.

95% that 2027 is the hottest year on record. 76% that world food prices average more than in 2026.

98% that a US state freezes new data centres. 91% that a leading AI lab publicly pauses training.

74% that Marine Le Pen wins the French presidency. 86% that her party’s candidate does.

56% that the average US mortgage rate ends 2027 above 7%.

14% that China blockades or quarantines Taiwan by the end of 2027. 10% that its forces land.

66% that US clothing stores outgrow the rest of retail.

14 Starship launches in 2027 as our middle estimate, in a range from 6 to 30.

96% that Avengers: Doomsday passes $1.5 billion worldwide by 31 March.

The reasons matter more than the numbers. So do the places where we part company with the magazine.

Washington and Paris

Trump: “will not be impeached” is too strong. The magazine expects the new Democratic House to impeach members of the cabinet and leave the president alone. We agree on the order. A cabinet impeachment gets 72%. Democratic leaders have already backed a case against Pete Hegseth, and their majority is large enough to lose a few votes. The likeliest way to miss is timing. A secretary who resigns or is fired before the vote no longer counts.

The Economist on Trump’s final chapter · The cabinet forecast

For Trump himself we have 38%. Party leaders have not committed to it. But investigations will run all year, and more than one chance in three is a long way from “will not”. Neither number means anyone leaves office. Impeachment is the charge. Removal takes two thirds of the Senate. Meanwhile we expect about 10 presidential vetoes in 2027, so the fighting will not be limited to hearings.

The Trump forecast · Presidential vetoes

France: we agree, with one caveat. Our number for Le Pen personally is 74%. She leads every runoff poll we looked at. Our number for a National Rally candidate, whoever it is, is 86%. Much of the gap is her unresolved court case. If she cannot run, the party can still win with someone else. The two forecasts were made a day apart, so read the gap as a direction, not a measurement.

The Economist on populism · Le Pen personally · Her party’s candidate

The magazine asks whether populism has peaked. Our answer depends on the country. The AfD has led Germany’s Politbarometer poll in twelve releases running, and we give it 76% to finish 2027 alone in first place. Reform UK has slipped to third in YouGov’s polling and gets 18% to end the year on top. That is three countries and three different tides. France votes in the spring.

The German poll · The British poll

Iran and Taiwan

Iran: the two endings are not equally likely. The magazine sketches a deal, then writes that the president “could just as plausibly swing the other way” and strike again. Our numbers are far apart. At least one US air or missile strike on Iranian territory in 2027 gets 88%. One limited strike is enough, and American forces carried one out on Qeshm island this month. A signed or jointly announced nuclear agreement gets 34%. Both can happen in the same year.

The Economist on Iran · US strikes · The nuclear agreement

For anyone who buys fuel, a third number matters more. Tanker traffic through the Strait of Hormuz gets 29% to be back to at least 80% of its 2025 level by December 2027. Ships need insurers and crews before a treaty helps them. Our middle estimate for Brent crude at the end of 2027 is $78 a barrel, against about $93 in mid-November. Cheaper, but not cheap.

Tanker traffic · Brent crude

Taiwan: pressure, not war. American commanders once named 2027 as the year China’s army should be ready to take Taiwan. The magazine notes that US intelligence does not expect an attack next year, and points to 2028 as the riskier stretch. Our numbers fit that. Chinese forces landing on or seizing any territory Taiwan controls by the end of 2027 gets 10%. Even a brief landing on a small island would count. An enforced blockade or quarantine gets 14%. A large, officially named military exercise around the island gets 74%. Expect alarming headlines, and ships that still sail.

The Economist on Taiwan · A landing · A blockade or quarantine · A named exercise

Food and mortgages

Heat: 95% is not quite “certain”. We are nearly as sure as the magazine, and for the same reason. A very strong El Niño is peaking now, and its heat reaches the atmosphere about three months later. That puts our number at 95%. The rest covers a Pacific that cools faster than expected or a large volcanic eruption.

The Economist on the hottest year · The temperature forecast

The number we would watch is about food. Later in 2027 the Pacific will probably flip to its cool phase, La Niña. We put that at 66%. The weather headlines will improve. Food prices will be slower. A harvest lost to drought stays lost, and the UN’s world food price index is already running above its 2026 average. We give 76% to a higher average in 2027 than in 2026. When the weather news turns good, check the price of bread.

La Niña · Food prices

Mortgages: waiting is a bet. “A new era of expensive money is here,” the magazine says. For a household it looks like this. On a $300,000 loan fixed for thirty years, the monthly payment of principal and interest is $1,799 at 6% and $1,996 at 7%. That is $197 more each month for the same house.

The Economist on bond markets · How a monthly payment is calculated

Freddie Mac’s average 30-year rate was 7.11% in mid-November. We give 56% to it still being above 7% in the last reading of 2027, and 82% to the Federal Reserve raising rates at least once next year. Our middle estimate for the ten-year Treasury yield at the end of 2027 is 5.0%. Anyone delaying a purchase until rates fall is taking close to even odds.

The mortgage forecast · Fed decisions · Treasury yield

AI and rockets

AI: nobody stops it, but many will try. The magazine says a halt in AI development is unlikely. We agree, and the interesting numbers sit just below a halt. A US state enacting a statewide moratorium on new data centres in 2027 gets 98%. New Hampshire and Maine are both working on one. One of the five leading labs publicly pausing training for safety reasons gets 91%. OpenAI and Anthropic each disclosed a pause in 2026.

The Economist on AI · State moratorium · Training pause

Neither slows what is already released. We give 76% to a freely downloadable model matching, by the end of 2027, the best model that exists on 1 January. The magazine makes a similar call. You can stop a crane or a training run. You cannot recall a download.

Open models

Space: fourteen launches, give or take a lot. “Space is about to get much more crowded,” the magazine writes. Starship sets the pace. It has been flying about six times a year. Our middle estimate for 2027 is 14 launches, with four chances in five that the count falls between 6 and 30. The range is wide because reuse is not routine yet. On its last flight SpaceX recovered the ship and lost the booster.

The Economist on space · The launch forecast

A sharper test is useful work. We give 76% to Starship deploying working Starlink satellites on at least five separate flights. In China, LandSpace gets 57% to fly a Zhuque-3 booster for a second time before 2028.

Starlink deployments · LandSpace reuse

Shops and cinemas

Ozempic: clothing wins, partly for a dull reason. The magazine predicts “revenge shopping”. People on weight-loss drugs slim down and replace their wardrobes. We give 66% to US clothing stores growing faster than total retail in 2027. Some of that has nothing to do with waistlines. Cheaper fuel shrinks the dollar sales of filling stations, which make up about a tenth of retail, and that flatters every other category.

The Economist on slimming drugs · Clothing sales

The drugs keep spreading. The health pollster KFF last found that 18% of US adults had tried one. We give 78% to a 2027 poll showing at least 20%. The drugmakers’ boom is cooling, though. Eli Lilly growing its revenue by another quarter gets only 15%.

The KFF poll · Lilly’s revenue

Hollywood: fatigue has limits. “Audiences are sick of movie franchises,” says the magazine’s headline, and studios make them anyway. The first test comes early. We give 96% to Avengers: Doomsday passing $1.5 billion worldwide by 31 March. Disney has already reported $50 million in advance sales.

The Economist on franchises · The Avengers forecast

Star Wars: Starfighter, due in May, looks ordinary beside it. Our middle estimate is $480 million worldwide by the end of August, in a range from $280 million to $860 million. Franchises are not dead. They are no longer automatic.

The Starfighter forecast

How we will score this

Each link above opens the forecast’s public page, with its exact rules, its reasoning and its sources. The first result arrives on 1 April, when we check the Avengers total. France votes in the spring. Most of the rest settle in January and February 2028.

We will publish every result then, misses included. The yes-or-no questions get a Brier score, which punishes a confident miss more than a cautious one. Where the magazine named an outcome that can be checked, we will score its call beside ours.

If you think one of these numbers is wrong, write down your own today. In a year we can both check.