U.S. ELECTRIC VEHICLES · FIRST-HALF 2026
Why EV electricity use can rise while new sales fall
Electricity demand comes from vehicles already on the road as well as new arrivals. A slowdown in new EV sales can therefore coexist with rising total electricity use, although fleet size alone does not explain every change.
EIA’s September 30, 2026 analysis reports an 8% increase in estimated U.S. light-duty EV electricity consumption in the first half of 2026 versus the second half of 2025. New EV sales fell 19% over the same comparison. This is a useful distinction for anyone following carmakers, charging businesses or electricity demand.
How far did the two measures move?
Electricity use index
100 → 108
A measure of energy consumed across the operating fleet.
New-sales index
100 → 81
A measure of vehicles newly sold during each half-year.
Open the complete chart data and index calculation
| Measure | Jul–Dec 2025 | Jan–Jun 2026 | Reported change |
|---|---|---|---|
| Estimated light-duty EV electricity use | 100 | 108 | +8% |
| New EV sales | 100 | 81 | −19% |
Index = 100 × (1 + reported percentage change ÷ 100). Electricity: 100 × 1.08 = 108. New sales: 100 × 0.81 = 81.
The inputs are rounded percentages published by EIA. The indices inherit that rounding. Dividing 108 by 81 would not produce electricity consumption per car: the electricity measure covers the operating fleet, while the sales measure covers new transactions.
What connects the showroom to the electricity meter?
A fleet accumulates over time. Selling fewer new cars than in the previous period can still add vehicles to that fleet. Meanwhile, existing vehicles continue to be driven. Retirements, movement between regions and vehicle usage also matter.
EIA’s model combines estimates of vehicles in use, electric miles traveled, efficiency and weather effects. It uses registrations where available and more recent sales information to extend lagged fleet estimates. That makes the electricity series more informative about fleet-wide energy use than a sales headline alone, but it also means the two series are not independent measurements.
For a charging operator, the practical question is narrower than national energy growth: how much charging happens at its sites, at what times, and at what margin? For a grid planner, geographic concentration and peak load matter. The aggregate estimate cannot answer either question on its own.
Which comparisons need care?
- Periods: the two headline changes use matching half-years. They are neither year-over-year nor seasonally adjusted growth estimates presented here. Different seasons and the half-years’ different lengths can affect totals.
- Coverage: EIA includes light-duty battery-electric and plug-in hybrid vehicles. Conventional hybrids and heavier EV categories are outside this electricity estimate.
- Measurement: the electricity figures are modeled estimates, not a national total directly read from dedicated vehicle meters. EIA’s methodology identifies potential errors in fleet estimates, driving patterns and efficiency.
- Revenue: higher electricity consumption is not a forecast of charging-company sales or profits. Charging location, tariffs, utilization and costs would need separate evidence.
Why the model’s revision process matters
Recent fleet estimates must bridge gaps in registration data. The documentation says monthly scrappage and interstate vehicle movement are not incorporated in the preliminary extension. Final registration and odometer inputs can arrive around 12–13 months after a reference year ends. Treat recent estimates as revisable, particularly when comparing individual states.
EIA also explains that utilities generally record vehicle charging within residential, commercial or industrial sales according to where charging occurs. Do not add estimated EV use to those electricity-sales totals as though it were an entirely separate category.
Does this show that lower sales no longer matter?
No. Slower additions can influence the fleet’s future growth. The current comparison simply shows why a change in new sales cannot be applied mechanically to electricity demand from all vehicles. A useful next check is whether fleet growth, electric miles per vehicle and charging location are moving together.
Can the figures establish the effect of an EV policy change?
They describe a period, rather than isolate a policy’s causal effect. A causal claim would need a comparison that addresses purchase timing, prices, model availability and other changing conditions. This article does not estimate that effect.
Sources, dates and reuse
Observation periods: July–December 2025 and January–June 2026. EIA analysis released September 30, 2026; supporting Monthly Energy Review is the September 2026 edition. Sources checked October 9, 2026 (UTC). This is analysis of those releases, not a new October data release.
- EIA, September 30, 2026: EV electricity use and new-sales comparison
- EIA, Electric Vehicle Consumption: model methods, limitations and data sources
- EIA, September 2026 Monthly Energy Review, Section 7, Note 4
- EIA copyrights and reuse policy
Original charts and calculations use selected statistics reported by EIA, with attribution. EIA permits reuse of its information products; privately contributed material remains protected. The underlying model uses third-party inputs, including S&P Global Mobility and Wards Intelligence. Their databases, graphics and records are not reproduced here; this article does not convey a license to them. No agency photographs or logos are used.
