Between 2005 and 2024, China more than doubled its oil consumption, accounting for over half of the global increase in oil demand during that period. But in 2024 Chinese annual oil demand declined for the first time in twenty years. This column develops a dynamic model which suggests that the rapid adoption of electric vehicles in China has been a major driver of this trend. In 2024 alone, EV diffusion displaced about 0.43 million barrels per day of gasoline, and this figure could quadruple by 2040 with an accelerated transition.
Between 2005 and 2024, China more than doubled its oil consumption, becoming the world’s largest crude importer and accounting for over half of the global increase in oil demand. After two decades of expansion, transport oil use began to level off after 2019 (Figure 1) and recorded its first annual decline in 2024, a turning point partly driven by the rapid diffusion of electric vehicles (EVs). This structural shift aligns with China’s broader strategy to boost its technological self-reliance and reduce external vulnerabilities (de Soyres and Moore 2024), amid a global reordering of energy and industrial supply chains.
Figure 1 World crude oil consumption by country
Figure 1 World crude oil consumption by country
Source: Statistical Review of World Energy 2025, Energy Institute.
As anticipated by Cherif et al. (2017), large-scale EV adoption can cut oil consumption by tens of millions of barrels per day within two decades, reshaping oil trade flows and influencing the global price of oil. Micro-evidence confirms that real-world energy savings from EVs depend on charging access and driving patterns (Bushnell et al. 2022, Rapson and Muehlegger 2023, Gessner et al. 2025). China stands at the centre of this transformation: through coordinated industrial policy, large-scale subsidies and tight integration between carmakers and battery producers, it has built the world’s largest EV market and manufacturing base (Ezell 2024, Bickenbach et al. 2024, DiPippo et al. 2022). This industrial strategy is a pillar of China’s geoeconomic positioning, with implications for firms’ competitiveness and global value chain dynamics (D’Orazio et al. 2024, Kee and Xie 2024, Gourinchas et al. 2024, Arezki and Van der Ploeg 2025). In 2024, EVs accounted for over 25% of new car sales in China and represented 11% of the total domestic vehicle stock, surpassed only by Norway in EVs penetration, with Chinese EVs accounting for about 60% of the global EV fleet (see Table 1 for vehicle definitions and Figure 2 for international comparison).
Table 1 Vehicles type and definitions
Table 1 Vehicles type and definitions
Source: IEA.
Figure 2 Electric passengers vehicles stock and growth by country
Figure 2 Electric passengers vehicles stock and growth by country
Source: CEIC; Our World in Data and IEA. Note: In the left panel, "Other" includes Diesel, Natural Gas, LPG and other alternative fuels. For the right figure the green bars show each country's share of electric vehicles (EVs) on its total vehicle fleet in 2024 while the orange bars indicate the growth in the number of circulating EVs from 2014 to 2024. In the right panel the blue bars show each country's share of the global EV fleet in 2024.
To assess the impact of EV adoption on Chinese gasoline consumption, in a recent paper (Bencivelli et al. 2025) we develop a dynamic model tracking the stocks of circulating passenger vehicles, distinguishing between internal combustion (ICE) and EV vehicles and their fuel consumption. The model (i) measures the EV-driven reductions in gasoline consumption between 2015 and 2024 by comparing actual Chinese gasoline consumption with the model-based counterfactual obtained under a "No EV adoption" scenario;1
(ii) projects gasoline demand up to 2040 under alternative EV-adoption paths; and (iii) compares long-run fuel savings by comparing these three scenarios to a "Constant EV adoption" baseline, holding the EV share in new car sales constant at 2024 levels.2
We use annual data on EV sales and the stock of ICE and EV vehicles in circulation and we follow Nguyen-Tien et al. (2025) to calibrate their respective scrappage rate.
Estimated gasoline consumption reduction between 2015 and 2024
Between 2015 and 2024, the adoption of EVs accelerated markedly in China, leading to a reduction in gasoline consumption. Figure 3 compares actual Chinese gasoline consumption in levels (solid blue line, left axis) with the "No EV adoption" counterfactual scenario (dashed blue line, left axis), and shows the difference between the two series (blue bars, right axis) in millions of barrels per day (mb/d).
Figure 3 Actual gasoline consumption and "No EV adoption" counterfactual
Sources: CEIC, IEA, Our World in Data. Notes: blue solid lines for realized data and blue dashed lines for the counterfactual of "No EV adoption". The blue vertical bars represent the difference between the actual and the counterfactual gasoline consumption in mb/d.
By 2020, EV adoption had already reduced gasoline consumption by 0.1 mb/d (3% of 2020 annual gasoline consumption) relative to the baseline. The effect grew steadily thereafter, reaching 0.43 mb/d in 2024 (12% of annual Chinese gasoline consumption in 2024). This reduction translates into an estimated 67.5 million tons of CO2 avoided emissions in 2024, roughly 0.6% of China’s total CO2 emissions in 2023, comparable in magnitude to the 1% drop in CO2 emissions observed during the "Zero-COVID" policy implementation.
Estimated gasoline consumption reduction, long-run projections for 2025-2040
We quantify the long-run effects of EV adoption in China relative to a baseline "constant adoption" scenario, which fixes the EVs share of new passenger car sales at its observed 2024 level of roughly 48%, projecting the evolution of China’s passenger vehicle fleet through 2040 under alternative EV adoption paths, linking these trajectories to gasoline demand and associated CO2 emissions.3
Figure 4 Scenarios on EV adoption and implications for annual Chinese gasoline demand, 2025-2040
Source: CEIC, IEA Our World in Data, OIES. Notes: Historical data are depicted in solid black lines.
We simulate alternative paths for EV adoption, measured as the share of EVs in new car sales, to assess how they shape China’s future gasoline demand. Slow, medium, and fast transitions imply progressively higher EV shares in new sales. A "no EV adoption" case is included purely as an extreme counterfactual.
Panel (a) of Figure 4 shows the resulting fleet compositions by 2040 under these scenarios. EVs account for roughly 60% of the total stock under the "fast adoption", approximately 50% under "medium" (and "slow") adoption and 30% under the "constant adoption" case. The tipping point where EVs outnumber ICE vehicles occurs in 2037 under fast adoption and in 2039 under the slow scenario; under the "constant adoption" baseline, ICE vehicles remain predominant even by 2040.
Panel (b) reports the corresponding gasoline demand trajectories. With "fast adoption", demand peaks as early as 2025, reflecting the rapid replacement of ICE vehicles. Under "medium adoption", the peak occurs in 2027, whereas in the slow transition case, gasoline demand continues to rise until 2028 before starting to decline. The gap between the oil consumption in the "fast adoption" case (green line in panel (b) of Figure 4) relative to the "constant adoption" baseline (grey line) quantifies the annual gasoline savings, which increase steadily over time, reaching about 1.0 million barrels per day (mb/d) in 2035 and 1.7 mb/d by 2040. As an example, the annual savings in 2040 would make an important contribution to global decarbonisation goals, corresponding to a reduction in emissions of roughly 267 million tons of CO2, around 2.3% of China’s total CO2 emissions in 2023.4
Conclusions
Our analysis suggests that the rapid adoption of EVs in China has been a major driver of the recent trend in the country’s oil consumption. In 2024 alone, EV diffusion displaced about 0.43 million barrels per day of gasoline. Our projections show that this figure could quadruple by 2040 with an accelerated transition. These findings carry profound implications that extend far beyond China’s borders. For global oil markets, they signal the weakening of one of the main sources of oil demand growth in the past two decades. For the global energy transition, they underscore how strong EV adoption could reshape China’s oil consumption patterns, delivering substantial emissions reductions.
Bencivelli L, A D’Orazio, S Emiliozzi and A G Gazzani (2025), "The Impact of China’s Electric Vehicle Adoption on Global Gasoline Demand", Bank of Italy Occasional Paper.
The "No EV adoption" scenario is a model-generated baseline used to calculate oil consumption savings, assuming the complete absence of passenger EVs in China between 2015 and 2024. This is constructed by setting the annual share of new EVs entering the market to zero for the entire period.
For the long-run scenarios we also report an alternative benchmark, the "No EV adoption" scenario, in which China is assumed never to have adopted EVs over the time span considered from 2015 onwards.
The model is calibrated to match the International Energy Agency’s Global EV Outlook (2025), Bloomberg NEF, and OIES projections, with the total passenger vehicle stock reaching about 495 million units by 2040. In 2024, China has approximately one passenger car for every five people, a much lower ownership rate than in advanced economies, where the ratio is close to one vehicle per 1.5–2 persons. Both the IEA and BNEF foresee a continued rise in vehicle ownership, driven by income growth and expanding household wealth. At the same time, UN demographic projections point to a decline in China’s population – from 1.42 billion in 2024 to 1.34 billion in 2040. Our long-term baseline incorporates both forces, assuming that by 2040 China will have about one car per three people, consistent with other international forecasts.
The gap between the fast-adoption scenario (green line in panel (b) of Figure 4) and the "no EV adoption" counterfactual (ochre line, representing a hypothetical China where only ICE vehicles remain in use) measures the implied annual oil savings. In this case, annual gasoline consumption savings are more than twice as large, reaching 3.7 mb/d by 2040. This corresponds to a reduction of 582 million tons of CO_2 for that year, equivalent to about 5% of China total annual CO_2 emissions in 2023.
