Electric cars have passed a tipping point in China, where they are over
half of new car sales and mostly cheaper than petrol cars before any
subsidy. Elsewhere, sales have fallen every time support was withdrawn.
Worldwide they are a quarter of new cars but one car in twenty on the
road, saving around 3% of the oil that road transport burns.
Electric cars have tipped in China, not yet worldwide
In China, yes; worldwide, not yet. A tipping point is when a change keeps going without being pushed, and the test a reader can hold is simple: when support ends, do sales hold?
In China they do. Electric cars are over half of all new car sales there, most now cost less than the petrol equivalent before any subsidy, and their share held when a purchase tax came back in 20261. A few European markets have begun to tip as well, and Norway tipped more than a decade ago78. But in the other big markets, sales have fallen wherever governments withdrew their support1.
The effect on oil use is real but still small. Electric vehicles of every kind saved about 1.7 million barrels of oil a day in 2025, as much as Indonesia uses in total1. That is around 3% of the oil burned on the world’s roads, a figure we derive from the International Energy Agency’s (IEA) own estimates of road-fuel demand4.
That's how much oil the world's electric vehicles saved in 2025, around 3% of what road transport burns
One new car in four, one car on the road in twenty
Electric cars won’t outnumber petrol and diesel cars on the road for a long time. BloombergNEF, a commercial forecaster, expects that to happen only in 2047, with more than a billion combustion cars still driving in 20406.
That is because a car lasts for years, so the fleet changes far more slowly than sales. More than 20 million electric cars were sold in 2025, a quarter of all new cars, up from under 1% ten years earlier12. But only about 75 million cars on the road are electric, 5% of the world’s total2. In China, where the switch started earliest, it is around 13%1. Oil savings follow the fleet, not the sales.
Oil saved: an Indonesia’s worth, 3% of road fuel
Electric cars saved about 1.2 million barrels of oil a day in 2025; counting scooters, buses and trucks, electric vehicles of every kind saved 1.7 million1. China accounts for about 1 million barrels a day of that, around 15% of what its road transport would otherwise have burned1.
In emissions, electric vehicles avoided a net 190 million tonnes of CO2-equivalent in 2025 (other greenhouse gases counted as the amount of CO2 that would cause the same warming), similar to all of Spain’s energy-related emissions1. That net figure already counts the power stations that charge them. Over its lifetime, a typical battery car causes more than 55% fewer greenhouse gas emissions than a petrol car, even allowing for how the electricity is made1.
And yet road transport’s own emissions are still rising. Tailpipe emissions grew almost 10% between 2015 and 2025, because people drove nearly 25% more kilometres1.
The total the world still emits, while road transport’s tailpipe emissions are still rising.
What a tipping point actually means
The popular definition comes from journalism. In 2022 Bloomberg’s Tom Randall observed that once electric cars reach about 5% of new sales, “early adopters are overtaken by mainstream demand”, and predicted that a quarter of US new car sales could be electric by the end of 202512. The US reached about 10%1. The 5% rule was a pattern across about 18 countries, never a tested cause12.
Researchers use a stricter test with four parts: electric cars’ share keeps rising, petrol-car sales stop bouncing back, prices reach parity, and carmakers shift their model ranges7. Applied to the data, it gives a split answer.
That finding assumes today’s policies stay in place, and it is deliberately cautious about buyers: it assumes they weigh the upfront price heavily against future fuel savings (a 20% discount rate, in the model’s terms), which makes the tipping point harder to reach, not easier7. Germany’s slump after its subsidies ended shows why the first assumption matters. Some of the same researchers have warned against wishful thinking about positive tipping points8.
China tipped on price; elsewhere, policy still does the pushing
Electric cars are popular in China because they are now usually the cheaper car. In 2025 nearly 70% of battery electric cars sold there cost less than their petrol equivalent before any government incentive, up from around 50% in 20211.
Price parity was bought. China’s government spent at least $230.9 billion supporting the industry between 2009 and 2023, mostly through buyer rebates and exemption from purchase tax, and that total leaves out newer local rebates, cheap land, power and credit, and state investment, which could be as large again11. Our reading is that fifteen years of guaranteed demand let carmakers build at a scale where the cars became cheaper to make; the sources document the money and the price, not the path between them.
The proof that China has tipped is what happened when support was cut back. A 5% purchase tax on electric cars returned in 2026, and electric car sales in China fell by more than 20% in the first quarter. But the whole car market fell about as much, and by April electric cars were over 60% of sales1. They have stayed above 60% for four months running5.
Outside China, the price gap is still open, and that is the other half of the story. In Germany, Italy and the UK, battery cars cost about 17% more than comparable petrol cars6. A subsidy closes that gap for the buyer; ending it opens the gap again. That is the likeliest reason sales outside China still rise and fall with policy, and the natural experiments bear it out.
When the subsidies stopped, sales fell
Every recent withdrawal of support in a large market outside China has been followed by a fall in sales. Germany ended its subsidies at the end of 20233, and the electric share of new cars fell from 24% to 20% in 20242. New Zealand removed its Clean Car Discount and sales dropped 70% in 2024. Canada ended its rebate and sales fell by more than 30% in 2025, from 17% of the market to 11%1.
The United States ended its federal tax credits after September 2025, and sales in the last quarter of the year were 45% lower than a year earlier; since then electric cars have held at 6 to 7% of sales1. Germany’s rebound to 30% in 2025 came with tighter EU limits on carmakers’ fleet emissions21: a different policy lever, but policy all the same.
The journalist who popularised the 5% rule made the same point at the time.
So the switch has had authors: China’s central government with its rebates and tax exemption, the EU with its fleet emission limits, and, on the way down, the governments that ended their subsidies.
Plug-in hybrids count, but often run on petrol
Yes, plug-in hybrids count: every headline figure, including the 25%, treats them as electric cars. A plug-in hybrid has a battery you charge from a socket and a petrol engine, and these cars make up about a third of electric car sales; the other 65% are fully battery-powered1.
In real driving, many run on petrol far more than their official figures assume. In Europe, where cars report their real fuel use, plug-in hybrids registered in 2021 emitted 3.5 times their official CO2 figure9. For 2023 models it was 4.6 times, leaving them only about 18% cleaner than ordinary petrol cars and hybrids10. The IEA’s oil savings estimate counts plug-in hybrids by how far they are assumed to drive on electricity1, and European data suggest those assumptions are too generous, so part of the 1.7 million barrels is probably overstated.
Global sales growth has stopped accelerating
Electric car sales are slowing in 2026. First-quarter sales were 8% lower than a year earlier, the first such fall since 20201, and by the end of August sales were up only 4% on 2025: Europe up 29%, North America down 21%, and China down 12%, with sales falling there while the electric share held, the same pattern as the first quarter5. Those are preliminary industry figures. The IEA’s forecast for the full year is 23 million cars, 28% of sales1, which would need a strong autumn.
The longer view is steady rather than accelerating. 2025 was the fifth year in a row that global sales grew by about the same amount, roughly 3.5 million cars1. That is growth along a straight line, which is not what you would expect of a change that feeds itself, and it is the core of the case against a global tipping point.
Tariffs and lithium could slow the price fall
Cheap cars tipped China, so whatever keeps cheap electric cars out of other markets slows the rest of the world. Mexico has put a 50% tariff on cars from China and other countries without a trade deal since January 20261. Our reading is that tariffs like this keep out exactly the kind of low-priced electric car that tipped China’s own market.
The cost curve is less certain than it looks too. Battery prices fell 8% in 2025, but lithium prices at the start of 2026 were more than twice as high as a year earlier, and the IEA warns that today’s record-low battery prices are partly propped up by losses among suppliers, which is unlikely to last1. If batteries stop getting cheaper, the price gap outside China closes more slowly, and sales there stay tied to policy for longer. BloombergNEF has already cut its long-term forecast for electric car adoption two years running6.
Three markets will show whether the rest of the world tips
The IEA expects petrol and diesel car sales never to return to their 2017 peak, in any of its scenarios1. If sales simply kept adding about 3.5 million a year, a rough extrapolation rather than a forecast, they would pass 37 million by 2030. The full 2026 figures, due in the IEA’s next outlook in 2027, will show whether this year met the 28% forecast.
Three markets will show whether the rest of the world tips the way China has: whether US sales recover without tax credits, whether Europe’s share keeps rising as its emission limits tighten, and whether China’s share holds as its purchase tax rises.
Even if all three go the right way, electric cars are winning new sales faster than they replace what is already on the road. On BloombergNEF’s forecast, more than a billion petrol and diesel cars will still be driving in 20406, so the oil savings will lag the sales share for years.