Skip to content

Design prototype — all content is real. Ten case studies are fully built, and three write-ups are still in preparation.

All work

UC Davis ITS · Hawai'i SB 2691 / HB 2030 · 2026 legislative session

A Self-Financing Clean Vehicle Rebate Fund for Hawai'i

A clean vehicle rebate program that asks the state treasury for nothing: $117.4 million and 22,519 rebates over five years, paid for by a one-time fee that half of Hawai'i's new car buyers never see.

Date
Feb 2026
Role
Co-author; fee structure and fund model
Methods
Feebate design, Revenue-neutral fund modeling, Fee incidence analysis
Tools
Excel/VBA

$117.4M

Fund created

Five-year fee revenue, with no general-fund appropriation

22,519

Rebates funded

Over five years; 20% reserved for low-income households

~50%

Buyers who pay nothing

New combustion sales fall below the $32,000 price threshold

$8,000

Largest low-income rebate

New battery-electric — $5,000 base plus the $3,000 top-up

Overview

Hawai'i imports every drop of fuel it burns. Road transport is 23% of the state's greenhouse gas emissions, households spend more than $2 billion a year on transport fuel, and a household at the high end spends $13,700 a year on gasoline — between 14% and 25% of its income, against a national average near $2,500. Switching that household to an electric vehicle saves roughly $5,300 a year and avoids 10.8 tonnes of CO₂. The barrier is the sticker price, and in September 2025 the federal tax credit that had been softening it expired.

That left a state with the strongest case for electrification in the country and no durable program to act on it — and no appetite for one, because a rebate program that draws on the general fund has to win an appropriation fight every session. So the question this work answers is not whether rebates help. It is whether a rebate program can be built that never asks for money.

The answer is a feebate. A one-time fee on the least efficient new combustion vehicles, charged only above a price threshold, fills a special fund that pays point-of-sale rebates for new and used electric and plug-in hybrid vehicles. The fund is sized to what it collects, so it closes every year at zero. The three-tier structure the 2026 bills carry — a fee scaled to fuel economy, levied only on vehicles above $32,000 — is the design set out in this brief, and the tier cutoffs and rates come from the fee model built alongside it.

The question

Every state that wants electric vehicles to reach ordinary buyers runs into the same wall: the incentive works, and the incentive costs money the legislature has to find. Hawai'i hit that wall harder than most. Its households are the most exposed in the country to imported fuel prices, its climate obligations are statutory, and its general fund is contested every session by schools, housing and hospitals. A program that has to be re-appropriated is a program that can be quietly not re-appropriated.

So the design constraint came before the design. Whatever the program did, it had to raise its own money, and it had to raise it without landing on the households it was meant to help — the objection that has killed vehicle charges in other states. That makes the important question distributional rather than fiscal: not how much does it raise, but who exactly pays, and does the answer survive being read out loud in a committee room.

The design

The fee is charged once, at purchase, on new internal combustion light-duty vehicles, and it is scaled to fuel economy as a proxy for emissions: 4% of sticker price below 22 combined MPG, 2% between 22 and 30, and 1% above 30. Nothing is charged on a used vehicle, and nothing on an electric or plug-in hybrid. Above that sits a price threshold — no vehicle selling below $32,000 pays anything at all, which exempts about half of new combustion sales outright.

The revenue funds point-of-sale rebates, which matter more than tax credits for the households in question because they arrive at the dealership rather than at the following April's filing. The fund carries a 5% administration cost and is sized each year to what it collects, so the balance closes at zero rather than accumulating a surplus a future session could sweep. The diagram below is the whole mechanism, including the part a proposal like this has to own: every rebate it issues removes one of its own future fee payers.

The proposal as a circuit. The fee flows in from one side of the new-vehicle market and out as rebates on the other, with no appropriation and no accumulating balance. The return arrow is the design's own limit: the fee base is combustion sales, and the rebates exist to shrink it.
New combustion vehiclepriced over $32,000Clean VehicleSpecial FundNew or used EV buyer,at the point of saleFEE · 4% / 2% / 1% OF MSRPREBATE · $1,000 – $8,000EVERY REBATE ISSUED REMOVES A FUTURE FEE PAYERNO GENERAL-FUND APPROPRIATION · 5% ADMINISTRATION · CLOSING BALANCE $0 EACH YEAR

The fee

Three tiers, set on combined MPG and charged as a share of sticker price. The pairing is what makes the structure work: in Hawai'i's 2025 sales the least efficient vehicles are also the most expensive, so a percentage-of-price fee sorts on emissions and on ability to pay at the same time, without needing a separate means test. Over 70% of 2025 new sales were rated below 30 MPG. Average MSRP is across all 2025 sales in each band; the share column counts only vehicles above the $32,000 threshold, which are the only ones charged.

The fee
Fee classCombined MPGFeeAverage MSRP, all 2025 salesShare of sales above thresholdFive-year revenue
Class 1Under 224% of MSRP$39,88244.0%$74.9M
Class 222–302% of MSRP$35,43240.6%$36.6M
Class 3Over 301% of MSRP$28,87115.4%$5.9M

Source: Ramji & Jamhar (2026), Clean Vehicle Rebate Program for Hawai'i, UC Davis ITS; Hawai'i 2025 new light-duty sales.

Figure 1

Who pays the fee, out of every 100 new gasoline or diesel vehicles sold in Hawai'i

Unit chart of 100 new gasoline or diesel vehicle sales in Hawai'i, divided by the fee each would pay. Roughly 50 of every 100 are priced below the $32,000 MSRP threshold and pay nothing. Of the remainder, 22 are rated under 22 MPG and pay 4% of MSRP, an average of $1,690; 20 are rated 22 to 30 MPG and pay 2%, an average of $894; and 8 are rated above 30 MPG and pay 1%, an average of $383. No electric or plug-in hybrid vehicle pays a fee, and no used vehicle pays a fee.

Source: Ramji & Jamhar (2026), Clean Vehicle Rebate Program for Hawai'i, UC Davis Institute of Transportation Studies. Fee shares from the Clean Vehicle Rebate Fund model, run on Hawai'i 2025 new light-duty sales by MSRP and combined MPG.

Data table
Who pays the fee, out of every 100 new gasoline or diesel vehicles sold in Hawai'i
Fee tierRateShare of new ICE salesAverage one-time fee
Below $32,000 — exemptNo fee49.9%$0
Under 22 MPG4% of MSRP22.0%$1,690
22–30 MPG2% of MSRP20.3%$894
Over 30 MPG1% of MSRP7.7%$383

The rebate

What the fund buys, at the point of sale. Twenty per cent of funded rebates each year are reserved for low-income households, who receive $3,000 above the base amount whether they are buying new or used — so the largest rebate in the program, $8,000 on a new battery-electric vehicle, is only available to the buyers least able to front the price. SB 2691 SD1 caps eligibility at $60,000 for a new vehicle and between $3,000 and $25,000 for a used one.

The rebate
VehicleBase rebateWith low-income top-up
New battery-electric$5,000$8,000
New plug-in hybrid$2,500$5,500
Used battery-electric$2,000$5,000
Used plug-in hybrid$1,000$4,000

Source: Ramji & Jamhar (2026); rebate amounts and price caps as carried in S.B. 2691 S.D. 1 (2026).

Figure 2

Clean Vehicle Rebate Fund revenue by fee class, US$ millions per year

Stacked bar chart of Clean Vehicle Rebate Fund revenue for each of five years, split by the three miles-per-gallon fee classes. Total revenue falls from $24.8 million in year one to $21.7 million in year five as combustion vehicle sales decline, giving $117.4 million over the period. Vehicles rated under 22 MPG contribute $74.9 million, or 64% of the fund, despite being 44% of the sales that pay a fee. Vehicles rated 22 to 30 MPG contribute $36.6 million, or 31%, and vehicles above 30 MPG contribute $5.9 million, or 5%. Spend equals revenue in every year, leaving a closing balance of zero.

Source: Ramji & Jamhar (2026), Clean Vehicle Rebate Program for Hawai'i, UC Davis Institute of Transportation Studies. Clean Vehicle Rebate Fund model, 50,000 annual light-duty sales with electric vehicle share rising from 15% to 30% across the five years.

Data table
Clean Vehicle Rebate Fund revenue by fee class, US$ millions per year
YearUnder 22 MPG · 4%22–30 MPG · 2%Over 30 MPG · 1%Total
Year 115.837.731.2524.81
Year 215.537.581.2324.34
Year 315.127.381.2023.70
Year 414.567.111.1522.82
Year 513.846.761.0921.69

Key findings

  1. 01

    Half of new car buyers never see the fee

    The $32,000 threshold exempts 49.9% of new combustion sales before any tier is applied. Among the half that are charged, the average one-time fee runs from $1,690 at the least efficient tier down to $383 at the most efficient. Nothing is charged on a used vehicle, which is where most lower-income purchases happen.

  2. 02

    The least efficient tier carries two-thirds of the fund

    Vehicles under 22 MPG are 44.0% of the sales that pay a fee but $74.9M of the $117.4M raised — 64% of the fund. That concentration is not a rate choice; it falls out of the fact that thirstier vehicles in Hawai'i also carry the higher sticker prices, averaging $39,882 against $28,871 above 30 MPG.

  3. 03

    The fund is designed to shrink

    Revenue falls from $24.8M in the first year to $21.7M in the fifth as combustion sales give way to electric ones — a 12.6% erosion of the fee base over five years, accelerating afterwards. This is the correct behaviour for a transition instrument and the wrong behaviour for a revenue source, and it is the reason the program is scoped as a bridge to 30% EV sales rather than as permanent policy.

  4. 04

    Equity is a budget line, not a preference

    Reserving 20% of rebates for low-income households is written into the fund's sizing rather than left to uptake: 4,504 of the 22,519 rebates, $33.2M of spend, of which $13.5M is the top-up itself. Because the fee exempts the used market and everything under $32,000 entirely, the households receiving that spend are largely not the households funding it.

What happened

S.B. 2691 passed Second Reading as amended, S.D. 1, on 12 February 2026 and was referred to Ways and Means, where it stopped. The House companion travelled further: H.B. 2030 crossed over as H.D. 2 carrying the same three-tier structure keyed to fuel economy, with the rates left blank — the placeholder the legislature uses when a number is still being negotiated — and was deferred by the Senate Transportation committee on 24 March 2026. The committee chair's stated reason was that Ways and Means "is not in the mood, at least for now, to do this". Neither measure advanced before the session adjourned, and the Hawai'i Electric Vehicle Association's own wrap-up records that none of the session's EV bills made it through.

What the deferral was not was a finding against the arithmetic. Nothing in the record disputes the fee base, its incidence or the fund's balance; the objection was appetite for a new charge on vehicles. That distinction is what makes the work reusable. The threshold, the tier cutoffs, the three rates, the rebate schedule and the reserved low-income share are all model inputs rather than assumptions welded into the structure, so the same design can be re-run against whatever numbers a later session is willing to carry — including the harder version of the question, which is what the program does after the fee base it depends on has done its job.

Prepared at the UC Davis Institute of Transportation Studies with Dr. Aditya Ramji. The fee structure, MPG tier cutoffs, price threshold and five-year fund projections shown here are from the policy brief and the Clean Vehicle Rebate Fund model built for it. Rebate amounts follow S.B. 2691 S.D. 1; the House companion, H.B. 2030, carried a slightly different schedule.