EV Policy by Country

Electric cars were 24% of new-car sales worldwide in the first half of 2026: above 60% in China, above 30% in the European Union, and about 7% in the United States (IEA definition, which includes plug-in hybrids). This page helps explain differences like these by cataloging, from primary sources, what each of ten major car markets actually does about electric vehicles: every purchase subsidy, company-car tax break, registration-tax difference, fleet standard, sales mandate and import tariff in force, each dated and linked to the document behind it, beside each market's own official battery-electric share of new sales.

The record itself is the interesting part. The United States ended all three federal purchase credits for vehicles acquired after September 30, 2025, and admits a Chinese-built EV only over a stacked 127.5% tariff. Norway, where battery-electric cars are 96% of new passenger cars, now charges electric cars a higher daily circulation tax than petrol cars (NOK 9.16 against 6.52) and has enacted the phase-out of its VAT exemption. And the largest policy differences are the least discussed: taxes are roughly half the pump price of gasoline across Europe, against one-seventh in the United States, and company-car and benefit-in-kind taxation, the category this inventory found hardest to verify anywhere, is excluded even from the IEA's own headline support metric.

What this page deliberately does not do: the shares carry their own denominators and are not directly comparable across markets, and observed differences in EV share cannot be separated into preference, policy, price, supply, product mix, geography and infrastructure with the available data. The page documents the record and leaves that judgment to the reader.

1. Cross-market IEA shares use one definition

The figures in this section alone use the IEA's harmonised term. Each is quoted from the Electric Car Markets in a Time of Uncertainty, July 2026, licensed CC BY 4.0 and licence-verified on 2026-08-31.

The denominator matters even within this harmonised series. Global electric-car sales were about 1% lower year on year in H1 2026, while total car sales were about 5% lower. Share rose because the whole market contracted faster. See global EV sales for the growth narrative.

2. Six kinds of EV policy in ten markets

Rows are alphabetical. A market enters only when all six categories are resolved from primary sources, including a sourced negative. Status, challenge and verification are separate fields. "Supranational" identifies an EU or EEA policy applying to the country; there is no European Union row.

U.S. subnational completeness: fees 51/51, purchase-incentive status 51/51, and ZEV status 51/51. The 13 ACC II state policies are named in the inventory data record: California Board Resolution 26-3; Colorado 5 CCR 1001-24 Regulation 20; Delaware 7 DE Admin. Code 1140; District of Columbia 20 DCMR chapter 9; Maryland Executive Order 01.01.2025.10; Massachusetts 310 CMR 7.40; New Jersey N.J.A.C. 7:27-29A; New Mexico 20.2.91 NMAC; New York 6 NYCRR Parts 200 and 218; Oregon's 2025-06-13 DEQ memorandum; Rhode Island 250-RICR-120-05-37; Vermont Executive Order 04-25; and Washington Ecology Publication 25-14-036.

3. Road-fuel taxation, July 2026

Tax made up 14.60% of the U.S. pump price of gasoline in July 2026, against 49.83% for the EU-27 weighted average, 51.46% in the United Kingdom and a range of 30.16% (Sweden) to 56.25% (Malta) across sixteen EU member states.

This is general road-fuel taxation, a tax on a substitute good. It is not an EV policy and does not appear in the table above. A higher pump price raises the gasoline expenditure a driver avoids by not buying gasoline; the tax component shown here is the part of the price that is policy. This table does not measure per-kilometre or net EV savings, which would require electricity prices and vehicle efficiency, and this section makes no claim about how fuel taxation affects EV shares.

This cross-market view is Europe-weighted: sixteen of its twenty-three markets are EU member states. China, Norway, Australia, New Zealand and Korea are absent because no tax-component source was verified for them.

Per-market notes: Brazil: Gross statutory content. A R$0.44/L gasolina-A subvention (MP 1.358/2026) offsets R$0.308 per blended litre; net share 29.30%. Price is the E30 blend. Canada: Our volume-weighted construction, band 17.5-17.8%; no official national tax average exists. Federal excise is $0.00 from 2026-04-20 to 2026-09-07 under Bill C-30, reverting to $0.10/L on 2026-09-08. India (Delhi): Delhi only; state VAT ranges from 1% (Andaman) to 35.2% (Telangana). Central excise INR 11.90/L after the 2026-03-27 cut, which PIB states was not passed to the pump. Japan: Broad measure incl. the 2.8 JPY/L petroleum and coal tax; excise-plus-consumption-tax-only is 25.98%. The provisional excise was abolished 2025-12-31 (53.8 to 28.7 JPY/L). Pump held near 170 JPY/L by a refiner subsidy, ~26 JPY/L effect at 2026-08-24. United Kingdom: Duty 52.95p/L unchanged since 2022-03-23; the scheduled 2026-09-01 step was cancelled 2026-05-20, holding through year end. VAT-registered fleet buyers reclaim VAT; their tax content is 34.79%. United States: Volume-weighted state average (2024 SEDS weights); EIA's own unweighted basis gives 13.27%. Federal 18.4 c/gal unchanged since 1993. Local option taxes excluded, so slightly understated.

All figures July 2026, verified 2026-09-01. Full sourcing, FX bases and the two-stage verification record are in the repository research file.

4. Dated policy changes and subsequent sales observations

This fixed corpus is non-exhaustive and non-random. Selection occurred after outcomes were already available, and the research sought subsidy removals, so unchanged markets are absent by construction. The rows are ungrouped and chronological. "Official" below describes the source series, not an estimate of policy effect.

The Denmark model result is deliberately narrow. Asplund, Jinkins, Lutz and Paizs estimate Tesla Model S registrations only: 1,965 actual over the study period versus 1,858 to 2,008 under their no-tax-change counterfactual. That model-specific timing result cannot be generalised to Danish EV demand, and this page draws no market-wide level-effect conclusion.

5. What different evidence can and cannot say about Chinese state support

State support was central to creating China's EV industry, supply chain and scale. The total of current support has not been quantified: CSIS and the OECD MAGIC database exclude the battery industry, and no source in the brief attempts a complete total. The evidence below therefore remains separated by object, denominator, period, scope, method and exclusions.

Operative legal finding. Regulation (EU) 2024/2754 is final and operative. Additional rates are BYD 17.0%, Geely 18.8%, SAIC 35.3%, Tesla Shanghai 7.8%, other cooperating producers 20.7% and all others 35.3%, on top of 10% MFN; the lesser-duty rule was not applied. It is under WTO challenge in DS630, with no panel report expected before Q2 2027, and six undecided General Court actions. Legal status does not settle the evidentiary interpretation.

Cost cross-check, not a control. IEA observations of about 30% lower ICE production cost versus Germany and about 35% lower BEV production cost versus advanced economies use different comparators and unmatched vehicle definitions. They are consistent with a large general China manufacturing-cost advantage; subtraction between them identifies no EV-specific contribution. Rhodium's firm comparison does not find counted direct grants and preferential financing dominant in that comparison, but cannot support an industry-wide verdict.

Export and price evidence. The identified vehicle VAT remission is 13% against a 13% levy, full remission rather than excess remission under Annex I(g); that says nothing about support under other theories. The battery export rebate was cut to 9%, then 6%, and is abolished from 2027-01-01. Rhodium measures BYD Seal U per-unit profit at about EUR 14,300 in the EU versus EUR 1,300 in China, while the IEA says exporters seek higher returns abroad. The EU proceeded ex officio under anti-subsidy law, not anti-dumping law. The exports-minus-overseas-sales gap is reported by the IEA as destination inventory, partly shipping lag, and is not evidence about zero-mileage exports.

Separate comparators and objections. The IEA reports that LG Energy Solution and Panasonic Energy would have had negative 2024 operating margins without U.S. production credits, while CATL margins were 10% to 15% in 2020 to 2024 and 18% in 2025. Those facts do not establish comparable scope or competitive effect. South Africa's general/MFN rate is 25% for BEVs and combustion cars alike; 18% is a preference available to qualifying EU/UK combustion vehicles, not the general combustion rate. Finally, a Beijing think tank critique makes a specific, testable objection to the OECD Loan Prime Rate benchmark. Its co-author is a former senior central-bank official; it is a serious objection, not a neutral audit.

6. United States series

The United States is included in the inventory because all three subnational lists reached complete named coverage. Its sales path remains a confounded descriptive break. The monthly values, definitions, source history and already-published descriptive analysis are on U.S. EV and hybrid sales; no memo-derived decomposition is repeated here.

7. Method, sources and vintage

Roster rule. A country enters only if all six cells resolve from a primary source. An explicit primary-source statement of nonexistence or absence from a complete official schedule can resolve a negative. Company-car treatment is not exempt. A subnational cell must have complete named coverage. The frozen rule produced ten rows, each with six resolved cells, verified 2026-08-31.

Cell schema. Every policy cell records value, summary, implementation status, challenge status, verification status, jurisdiction level, announcement/effective/expiry dates, primary source and verification date. Every share records powertrain and vehicle scope, period, numerator, denominator, source URL and retrieval basis. Tense follows the structured statuses: Canada's mandate repeal and the U.S. CAFE reset are proposed; the EU CO2 revision remains in co-decision; the 2026-08-31 NHTSA document is interpretive; Canada's quota is in force.

Fuel taxation, added 2026-09-01. Section 3 was originally omitted because the tax component could not be isolated outside Europe. A dedicated verification pass established retail prices and statutory tax rates from primary sources for all six non-European markets, and both tax bases from the Oil Bulletin for sixteen EU member states, for the common month of July 2026, meeting the plan's Gate A, and the section was added. Non-European tax-exclusive values are calculated residuals, labelled as such in the section.

Excluded markets and gaps. Twenty-eight scoped markets were excluded because at least one of the six instrument cells lacked a primary-source resolution. The briefs do not contain a final per-market exclusion-cell ledger, so this page does not reconstruct or guess one. Company-car or benefit-in-kind treatment was the dominant documented coverage gap. The released page therefore covers 10 of 38 scoped markets and excludes 28. It is not a maintained world policy tracker. Blocked or automation-resistant hosts recorded in the briefs included acea.auto, regjeringen.no, retsinformation.dk, ft.dk, canada.ca, dep.nj.gov, tax.colorado.gov, svenskforfattningssamling.se, economie.gouv.fr, VicRoads, SRO Victoria, austlii.edu.au and several U.S. state-code portals.

Publication-day refresh gate. Before deployment, recheck the EU duty table and undertakings; DS630 and the six General Court actions; Canada's quota; Türkiye's implementation; the UK ZEV consultation; the EU CO2 proposal; Canada's proposed mandate repeal; the U.S. CAFE proposal; China's battery export rebate; Singapore's EEAI; Germany's PHEV window; and Norway's signalled 2027 VAT change. These page values remain stamped 2026-08-31 until that human gate is run.

Source boundary. Policy cells come from the committed verification briefs, with the plan's corrections overriding earlier conflicts. Country shares come from KBA, SDES, CBS, SCB, Statistics Denmark, SSB, DfT, Statistics Canada, the NVES Regulator and the site's Argonne-backed SQLite series. IEA figures appear only from the licence-cleared July 2026 report. The Policy Explorer PDF and downloadable GEVO dataset are not reproduced.