The Autumn Budget Audit: Deconstructing eVED Volumetric Mileage Traps vs. Fleet BiK Roadmaps

From Fiscal Protection to Treasury Normalisation

For most of the last decade the tax treatment of the battery electric vehicle (BEV) in the UK has been an exercise in deliberate asymmetry. Vehicle Excise Duty (VED) was set at zero for zero-emission cars, Benefit-in-Kind (BiK) was held at 2% for years, and, because electricity carries no equivalent of fuel duty, the marginal cost of a mile driven on a charger was untouched by any per-mile levy. The Autumn Budget 2026 is the point at which that arrangement is formally scheduled to end. It does not end abruptly. The measures are staged, with immediate stability for charging infrastructure and company car taxation, and a structural per-mile charge, the Electric Vehicle Excise Duty (eVED), set to commence in April 2028.

This article takes the measures as they were set out in the Chancellor's statement and deconstructs them as an engineering and financial-modelling problem. It first quantifies the volumetric mathematics of a flat per-mile charge, then audits the telematics architecture that the Treasury intends to use to collect it, then follows the Benefit-in-Kind trajectory and the revised Approved Mileage Allowance Payment (AMAP) rates through the fleet accounts, and finally sets the targeted £100m public charging fund against the extended fossil-fuel duty freeze. The intended readers are fleet operators, low-carbon transport consultants and automotive financial modellers, so the emphasis is on the arithmetic that drives a procurement decision rather than on the political framing.

A note on sources and status is needed at the outset. The figures below (3p per mile for BEVs, 1.5p per mile for plug-in hybrids, the 2030/31 BiK rates, the revised AMAP rates of 55p and 35p, the £100m fund and the £150 annual saving) are taken from the Budget as briefed to this publication. Several of them depend on legislation and consultation outcomes that are not yet complete, and a model built on them should carry a clear assumptions register that is revisited when the Finance Bill text and the eVED consultation response are published. Where this article derives a number, the derivation is shown so that the reader can substitute the final figures.

The Volumetric Physics of a 3p-per-Mile Tax

A per-mile charge differs from every other motoring tax in the UK in one important respect: its incidence scales linearly with distance and is indifferent to the efficiency of the vehicle. Fuel duty, by contrast, is a tax on the volume of energy purchased, so a more efficient internal combustion engine (ICE) car pays less per mile. VED is a fixed annual charge, so it falls more heavily on the low-mileage driver. The eVED is a pure distance tax, and the arithmetic is therefore simple enough to be stated in one line: annual liability equals the rate in pence per mile multiplied by the annual mileage.

For the standard profile used in most household-level modelling, 8,500 miles per year, a 3p per mile rate produces 8,500 × £0.03 = £255 per year. For a plug-in hybrid (PHEV) at 1.5p per mile, the same mileage produces 8,500 × £0.015 = £127.50 per year. These headline values are modest in isolation. The modelling significance lies in how the charge scales with the use profile, because the fleet operator and the high-mileage private driver sit far above the household average.

1. 8,500 miles per year: BEV £255; PHEV £127.50.
2. 12,000 miles per year (a typical user-chooser company car): BEV £360; PHEV £180.
3. 20,000 miles per year (a regional sales or service vehicle): BEV £600; PHEV £300.
4. 30,000 miles per year (a high-mileage fleet asset): BEV £900; PHEV £450.

Over a five-year hold period the linear scaling matters more than the single-year figure. At 8,500 miles per year, five full years of the charge would amount to £1,275. Because the charge starts in April 2028, a vehicle acquired in late 2026 would not bear it for its whole first five years. A vehicle taken on in November 2026 and disposed of in November 2031 would bear the charge for approximately 3.6 years, or about £920 at 8,500 miles per year. The modeller must therefore align the start date of the charge with the vehicle's hold window rather than applying the annual figure to every year of a notional five-year cycle. A vehicle ordered in 2028 or later will bear the charge for its whole life.

The Per-Mile Cost in Context

To judge the significance of 3p per mile, it has to be set against the energy cost of the mile itself. A typical BEV consumes between 3.0 and 4.0 miles per kWh in mixed use. At 3.5 miles per kWh, 8,500 miles requires about 2,430 kWh of delivered energy. At a standard-variable home electricity price of 25p per kWh, that is roughly £607 per year, or 7.1p per mile. On that basis the eVED adds £255 to £607, an increase of 42% in the energy-plus-tax cost of the mile. On a smart off-peak tariff of 7.5p per kWh the energy cost of the same mileage falls to about £182, and the £255 of eVED then exceeds the energy bill itself by 40%. At the other extreme, a motorway rapid charge at 70p per kWh costs 20p per mile, and the 3p charge adds 15%.

The comparison with a petrol car makes the policy logic visible. Fuel duty at the main rate of 52.95p per litre, on a car returning 40 miles per gallon (4.546 litres per gallon), equates to 52.95 × 4.546 / 40 = 6.0p per mile before VAT is applied to the duty-inclusive price, and about 7.2p per mile after it. At 8,500 miles per year the petrol driver therefore contributes in the order of £510 in duty, or £610 including the VAT charged on it. The eVED at 3p per mile reaches roughly half of that contribution. Policy commentary frequently frames this as a transitional parity target, with the BEV driver moving from zero towards, but not reaching, the duty paid by an equivalent ICE driver. The fleet modeller should treat the 3p rate as a starting value and include a sensitivity for inflation indexation, because a flat nominal rate that is uprated annually will erode the cost advantage of the BEV over a long hold period.

The PHEV at 1.5p per mile occupies a favourable position. It pays half the BEV rate for the same mileage despite consuming petrol, on which it also pays fuel duty. The intention is evidently to recognise that the PHEV is already contributing to the Exchequer through its liquid fuel use, but the effect is that a PHEV driver who operates predominantly on battery receives a lighter eVED than a BEV driver with an identical electricity profile. A fleet that has been using PHEVs as a low-BiK compromise should model the actual electric-running share, because the eVED, unlike the BiK system, charges for the mile and not for the declared emissions figure.

Telematics and Data Architecture of eVED Connect

A pay-per-mile tax is only as robust as the mileage record on which it is calculated, and the Treasury's proposed collection mechanism, referred to as eVED Connect, is a data-engineering problem before it is a fiscal one. The design intent is to exploit the cellular connectivity already embedded in the majority of new vehicles, so that odometer readings can be reported without the driver or fleet administrator having to submit them manually. In outline, the architecture has four layers: a data source in the vehicle, a transport layer, a processing layer within government and a reconciliation layer between the estimated and the actual liability.

Data Source and Transport

The authoritative quantity is the odometer reading, which in a modern vehicle is held in a control unit and is exposed through the manufacturer's connected-car back end rather than through a standard open interface. A telematics control unit (TCU) with an embedded SIM transmits the vehicle's state, including odometer value, at intervals defined by the manufacturer's software. That reading is available to the manufacturer's cloud platform and, subject to consent and agreement, to third parties. The Treasury's model, as described, relies on this existing flow, which is attractive because it requires no new hardware in a new car. However, it creates a dependency on dozens of manufacturer back ends, each with its own API, data semantics and update cadence. Normalising those into a single mileage ledger is a non-trivial integration task, closely analogous to the problem faced by open-banking aggregators.

The coverage problem is the second constraint. Vehicles without an active connectivity service, older vehicles in which the TCU is not enabled, vehicles used in poor coverage areas and vehicles on which a subscription has lapsed will not report. The framework therefore needs a fallback, which is the second route described in the Budget: an upfront annual estimate. In that mode the keeper declares or is assigned an expected mileage, pays the corresponding charge in advance, and receives a reconciliation when the true figure is established at an annual odometer check, typically at the MOT. The MOT is a convenient anchor because the test already records the odometer reading in a government database, which gives a verifiable data point at no additional cost to the keeper.

Processing, Reconciliation and Refunds

The processing layer must perform three functions. It must ingest and validate the mileage stream, flagging anomalies such as odometer rollback, implausible daily jumps and gaps in reporting. It must compute the liability using the rate band that applies to the vehicle, which means distinguishing BEVs from PHEVs and, presumably, taking account of any exemptions. And it must reconcile the paid-in-advance estimate against actual mileage, producing either an additional charge or an automated refund. The Budget's commitment to automated refunds is significant, because a mismatch between estimated and actual mileage will be the normal state of affairs and not an exception. If the system overestimates mileage for a low-mileage keeper, the Exchequer holds the keeper's cash interest-free until reconciliation, and an efficient refund process becomes a matter of fairness as well as administration.

From a fleet operator's viewpoint the more consequential issues are administrative. A fleet of several hundred vehicles on mixed contracts, with some on employee-owned vehicles under a grey fleet arrangement, needs to know which entity is the liable keeper for each vehicle at each date, how mileage is attributed when a vehicle moves between depots or between drivers, and how a short-term hire or a pool vehicle is handled. The difference between the registered keeper and the operational user is a long-standing source of friction in the VED and the penalty charge regimes, and the same ambiguity will apply here. The operator should assume that its fleet management system will need to produce an auditable record, reconcilable to the Treasury's ledger, of which vehicle was held by which entity for which period.

Privacy and Compliance Boundaries

The odometer reading is a low-resolution data point compared with a location trace, and the Treasury has been explicit that the collection is concerned with distance and not with where the distance was travelled. That distinction is important legally and politically, because a mileage figure is considerably less intrusive than a journey history. Nevertheless, the data protection analysis is not trivial. A mileage record linked to a vehicle registration is personal data when the keeper is an individual. The controller, the lawful basis, the retention period, the rights of access and erasure, and the transfer arrangements with manufacturers who process the data on the government's behalf all have to be defined. A data protection impact assessment is the expected instrument.

The practical compliance friction for operators arises in three places. First, a manufacturer's connected-car consent is often given by the driver in the vehicle's infotainment system, and a fleet operator may find that the consent chain does not allow automatic sharing with the Treasury without a further step. Second, a fleet that has deliberately disabled the vehicle's connectivity for privacy or cost reasons will find that the default reporting route is unavailable to it. Third, if the framework incorporates a standard for data accuracy, then a fault in the TCU that causes a gap in reporting could lead to an estimated charge being raised in place of the actual one. A fleet should therefore build monitoring of the reporting status of each vehicle into its operational controls, just as it monitors MOT and tax status today.

Fleet BiK Roadmaps and AMAP Mileage Accounting

If the eVED is the long-term overhead, the Benefit-in-Kind regime is the immediate determinant of the cost of a company car for the employee, and the Budget has provided the fleet sector with something it has requested for years: a published trajectory with enough lead time to plan a lease cycle. The zero-emission band has been rising in steps, and the legislated path reaches 7% in 2028/29 and 9% in 2029/30. The 2030/31 rate is set at 10%, which completes what is often described as a slow crawl from the 2% of earlier years to a long-term position that remains well below the rates charged on petrol and diesel cars.

Worked BiK Exposure

The employee's annual tax is the product of the car's P11D value, the appropriate percentage and the employee's marginal rate. Take a BEV with a P11D value of £45,000 and a higher-rate taxpayer at 40%.

1. At 3%: 45,000 × 0.03 × 0.40 = £540 per year.
2. At 5%: 45,000 × 0.05 × 0.40 = £900 per year.
3. At 9%: 45,000 × 0.09 × 0.40 = £1,620 per year.
4. At 10%: 45,000 × 0.10 × 0.40 = £1,800 per year, or £2,025 for an additional-rate taxpayer at 45%.

The figure to carry from this table is the rate of change. Over the period from 3% to 10% the employee's annual charge on the same vehicle more than triples, an increase of £1,260 for the higher-rate taxpayer. That is a larger sum than the 2028 eVED on a typical mileage, and, unlike the eVED, the BiK charge does not depend on the distance driven. For the salary sacrifice scheme the comparison is further affected by the employer's National Insurance contributions, because the employer's saving on the sacrificed salary is partly offset by the rising BiK, and the net advantage of the scheme narrows as the percentage climbs. A model of the scheme's attractiveness to employees should be rerun at each step of the trajectory and not assumed to be stable.

The Expensive Car Supplement (ECS) adds a further dimension for higher-value vehicles. The supplement applies an additional annual charge to cars with a list price above a threshold, and it applies to zero-emission cars as it does to others. The threshold and the supplement are subject to annual change, so the fleet specifier should check the current values at the time of order, but the structural point holds: a premium BEV with a high list price is exposed both to the rising percentage and to the supplement, and the real cost of the company car for the employee can therefore diverge sharply across models with a similar range and capability.

The First AMAP Update Since 2011

The second corporate item is the revision of the Approved Mileage Allowance Payment rates for employees using their own vehicles for business. These rates have been fixed at 45p per mile for the first 10,000 business miles in a tax year and 25p per mile thereafter since 2011, an interval over which the cost of motoring has risen substantially. The Budget, as briefed, sets the standard rate at 55p per mile and the rate above the 10,000-mile threshold at 35p per mile.

The effect on an employee who drives 12,000 business miles in a year can be calculated directly. Under the old rates the tax-free allowance is 10,000 × £0.45 + 2,000 × £0.25 = £4,500 + £500 = £5,000. Under the revised rates it becomes 10,000 × £0.55 + 2,000 × £0.35 = £5,500 + £700 = £6,200, an increase of £1,200 or 24%. For a grey fleet of 200 drivers at that mileage, a policy of paying at the approved rate would raise the annual reimbursement budget by £240,000. Companies are not obliged to pay at the approved rate, but a rate below it makes the employee worse off relative to the allowable figure, and a rate above it creates a taxable benefit. The consequence is that the approved rate acts as a reference point for employee expectations even where it is not binding.

The AMAP update interacts with the eVED in an unexpected way. When the eVED begins, the employee who drives a BEV for business will be paying 3p per mile in a tax that is outside the reimbursement formula, while the AMAP rate remains a single figure for the vehicle class. A company that reimburses at 55p per mile is not obliged to track the eVED, but the allowance in effect covers it. For the employee on a lower mileage this provides a margin; for the one on a high mileage, with the 35p band, the margin is thinner. There is also a separate advisory electric rate for the reimbursement of home charging of company cars, set by HMRC, and the fleet should confirm whether it is reviewed in step with the AMAP update.

Public Charging Subsidies and the Fuel Duty Asymmetry

The third pillar of the Budget is concerned with who can charge cheaply at all. The distinction between the driver with a driveway and the driver without one is the largest single determinant of the cost of electric motoring in the UK, because domestic electricity on a home tariff costs a fraction of public charging. The Budget therefore allocates £100m to a public charging fund focused on lower-powered, on-street infrastructure, with the stated aim of reducing charging rates by around £150 per year for motorists who cannot charge at home.

Modelling the £150 Intervention

A saving of £150 per year can be translated into a unit price reduction. For a driver of 8,500 miles per year at 3.5 miles per kWh, the annual energy requirement is about 2,430 kWh. If the whole of that energy were purchased on-street, £150 equates to a reduction of 6.2p per kWh. A typical on-street lamp-post charger tariff is in the range of 30p to 50p per kWh, so a reduction of that order would be material, bringing the effective price down by between an eighth and a fifth. If the on-street driver purchases only a share of their energy from on-street sources and the rest from rapid chargers, the per-kWh reduction needed to deliver the same annual saving would be larger. The headline of £150 is therefore best read as an average outcome for a defined user profile and not as a guaranteed unit rate.

The scale of the fund can be tested against the target population. Dividing £100m by £150 gives approximately 667,000 driver-years of saving. If the fund is spent over several years, the number of drivers supported at any one time is correspondingly lower, and the funding mechanism (capital grants to local authorities and operators, or revenue support per kWh dispensed) will determine whether the saving appears at the charger or only as a reduction in the cost of installation. A capital subsidy only reduces the unit price if the operator passes it on, and the fleet adviser should expect the retail effect to depend on the competitive structure of the local market.

The Fuel Duty Asymmetry

Alongside this targeted measure sits the extension of the 5p per litre fuel duty freeze to July 2027. In macroeconomic terms the two policies pull in different directions. The charging fund lowers the cost of one specific segment of electric motoring by around £150. The duty freeze maintains a reduction that, for the petrol car at 40 miles per gallon and 8,500 miles per year, consumes about 966 litres, and therefore saves approximately 966 × £0.05 = £48 per year relative to the unfrozen position, or about £58 including VAT. The sums are small per vehicle but very large in aggregate across the national fleet, and their fiscal cost is a function of the whole petrol and diesel car park rather than of the BEV population.

The distortion arises because the price signal to switch from ICE to BEV is weakened by each extension of the freeze, while the same Budget announces a tax that will remove the BEV's zero running-tax status. The consultant advising a fleet on the pace of transition should regard the freeze as a temporary protection of the ICE running-cost position, not a permanent one, and should model the post-July 2027 fuel duty path explicitly. If the freeze ends, the duty on petrol rises, and the relative TCO of the BEV improves even after the eVED. If it is extended repeatedly, the BEV advantage narrows.

Comparative Matrix: Fiscal Position of the EV Owner Across the Budget Horizon

Two features of the matrix deserve emphasis. The first is that the eVED affects every BEV owner by the same absolute amount per mile, but the proportionate effect on the cost of the mile is strongly dependent on the cost of the energy. The driver with the cheapest energy is hit hardest in relative terms, and the driver with the most expensive energy is hit least. This is the opposite of the pattern for the charging subsidy, which is aimed at the most expensive energy segment. The second is that the fleet user faces the largest number of simultaneous changes, since the BiK, the AMAP revision, the eVED and the data compliance requirements all converge on the same asset within a four-year window.

Recalibrating the Five-Year TCO Model

The practical consequence of the Budget is that a fleet TCO model constructed in 2025 or early 2026, in which the BEV running costs consist of energy, maintenance and insurance, is now incomplete. The following changes should be made to a model before it is used to support an order placed in the current cycle.

1. Add a dated eVED line. The charge should start in April 2028 and be calculated as mileage multiplied by the rate, with a switch for the PHEV rate. A sensitivity for annual uprating, for example in line with the Consumer Prices Index, should be provided, because a nominal rate that is indexed will rise by about 2% to 3% per year.
2. Separate the keeper from the user. The model should state which party bears the eVED, and whether it is recharged to the driver under the vehicle policy, since the answer differs between a leased pool vehicle and a salary sacrifice car.
3. Index the BiK path by tax year. Each year of the hold period should use the percentage for that year (4% in 2026/27, 5% in 2027/28, 7% in 2028/29, 9% in 2029/30 and 10% in 2030/31, subject to confirmation against the final legislation), and the model should show the employee's net monthly cost for each year so that the effect on take-up is visible.
4. Reprice grey fleet reimbursement. The new AMAP rates should be applied to the actual mileage distribution of claimants, with the threshold at 10,000 miles built into the calculation, and the result compared with any lower rate in the current policy.
5. Model the charging mix explicitly. The split between home, depot, on-street and rapid charging determines the baseline against which the eVED is a larger or smaller proportion, and determines whether the on-street fund is relevant to the employee population.
6. Model the fuel duty path for the comparator. The ICE or hybrid alternative should use a duty assumption beyond July 2027 that is stated and defensible, not an implicit continuation of the freeze.

Residual Value and Procurement Implications

The eVED will be borne by whoever holds the vehicle, and that raises the question of how it affects used-market pricing. A used BEV that will incur the charge for many years has a slightly lower operating-cost advantage than it would have had without it, and a buyer who is rational in a TCO sense will pay less. If the charge is £255 per year at 8,500 miles per year, and the buyer's horizon is four years at a discount rate of 5%, the present value of the charge is about £900, and that is the order of magnitude by which the used-market price could adjust, all other things being equal. In practice the effect on residual values will be muted by other factors, such as battery health reporting, supply of new vehicles and the manufacturers' pricing, but a lessor setting residual values for a contract that ends after 2028 should include the effect in its stress cases.

For procurement, the practical guidance is to seek flexibility. Contracts that contain a pass-through clause for changes in taxation allow the lessor to recover the eVED from the lessee, and the operator should read those clauses carefully before signing, since an uncapped pass-through exposes it to the full amount of any future rate rise. Operators with large fleets should also ask lessors to confirm how mileage data will be obtained for eVED reconciliation, and whether the lessor or the lessee will administer the estimate and refund process. These are the questions that will determine the operational cost of the tax even more than the rate itself.

Engineering Conclusions

The Autumn Budget 2026 is best read as the beginning of a managed convergence between the tax treatment of electric and liquid-fuelled motoring. The measures are individually modest. A 3p per mile charge adds £255 per year at an average mileage. A BiK increase from 3% to 10% adds some £1,260 per year to the higher-rate employee's tax on a £45,000 car. A revised AMAP adds 24% to the grey fleet reimbursement at 12,000 miles. Taken together, however, they remove the assumption of a permanently zero-tax running profile on which many early TCO models were based.

What remains in favour of the BEV is substantial. Even with the eVED, the tax per mile remains well under half of the fuel duty and VAT paid by an equivalent petrol car, the energy cost per mile remains lower for any driver with access to a reasonable domestic or depot tariff, and the BiK rate remains well below the rates charged on ICE vehicles. The charging fund, if it reaches the retail price, narrows the gap for the driver without a driveway. The decisions that are now more sensitive than before are the choice of charging route, the choice of vehicle value relative to the Expensive Car Supplement threshold, the pass-through terms of the lease and the quality of the mileage data on which the tax will depend.

The recommendation to the specifier is to rebuild the model with dated, switchable policy layers, so that each element (eVED rate, BiK percentage, AMAP rate, fuel duty path, charging subsidy) can be changed independently when the final legislation and consultation responses are published. A model that is structured in this way remains valid when the details change, and a decision made on it can be defended on the basis of the evidence that was available at the time.