Totally agree the tax code is a blunt instrument. The last 18 months have been a natural experiment: same cars, different eligibility switches, dealer POS “transferability” turning on, and a lot of noise in money factors. A few datapoints and practical tweaks from what I’ve seen or dug into:
Dealer/OEM capture: Rigorous public data is thin, but there are decent ways to see it. Around eligibility flips (e.g., when certain trims lost/gained 30D in Jan 2024), advertised discounts and doc/add-on fees moved in suspiciously convenient ways. Not universal, but common enough. If someone has access to PIN or TrueCar transaction series, a difference-in-differences on MSRP-capped vs not-capped models around rule changes would quantify capture. My back-of-envelope scrape on two crossovers showed advertised discounts shrinking by ~$1,500 the week POS credits turned on, then drifting back after inventory built. Not proof, but the pattern’s consistent with partial capture.
Leasing to access credits: The effective value of the “commercial” credit depends on how hard the finance arm leans on the money factor and residual. A lot of leases showed MFs in the 0.0020-0.0030 range last year (roughly 4.8-7.2% APR) when buy rates were lower, plus $650-$1,095 acquisition and $350-$595 disposition. If the lessor inflated residuals to keep monthly low, you pay at turn-in or via mileage penalties. On two quotes I compared head-to-head, the $7,500 headline passed through looked more like $4,000-$5,000 net after finance charges and fees. You can sanity-check quickly: monthly delta vs otherwise identical purchase, multiply by term, subtract fees you wouldn’t pay on a purchase, and compare against $7,500.
POS transferability and low-income buyers: One nuance that gets missed-starting 2024, the point-of-sale transfer means folks with little or no tax liability can receive the full amount so long as they meet the other rules; Treasury pays the dealer and doesn’t claw back due to insufficient liability. That’s an improvement. But income caps + MSRP caps still filter out a lot of low-cost real choices and push people toward leases or specific trims.
Used-EV incentives and supply games: Where states have used credits or sales tax exemptions (NJ, WA previously, OR’s Charge Ahead, CO), we’ve already seen programs add hold-period rules and one-per-X-years caps because of quick-flip behavior. The fix seems straightforward: minimum 12-month ownership before another claim, VIN-level tracking, and a “first transfer only” rule that’s enforced at DMV with odometer attestations. Programs that lacked those had to pause funding after burns.
Grid-aligned payments actually work: Utilities that pay per off-peak kWh or run managed charging pilots are the closest thing we have to outcome-based incentives today. Con Ed’s SmartCharge NY has paid roughly $0.10/kWh off-peak plus seasonal bonuses. Xcel’s Optimize EV, SMUD, and a bunch of Midwestern IOUs pay $5-$20/month for enrollment plus off-peak bonuses. Participation rates jump if you make it set-and-forget and pay monthly. That’s the model to scale: it drives the behavior you actually want and it’s hard for dealers to skim because it bypasses them entirely.
Battery health and the used market: Any “battery health dividend” needs a standard. Right now you’ve got third-party scorecards (Recurrent, etc.) and OEM BMS readouts that aren’t harmonized. The EU battery passport requirement will force SoH disclosure over there. If DOT/EPA piggyback on that concept-secured SoH via a common interface-you can safely tie a modest annual credit to verifiable capacity after year 5. That would immediately help valuation transparency and tamp down FUD in the used market.
Refundable credits without dealer capture: Pay them at DMV registration instead of at the showroom. You verify income/eligibility online, upload the signed buyer’s order, DMV pays you directly once the VIN is registered in your name and the OTD price is logged. Dealers can’t “adjust” a rebate the buyer receives outside the deal. If you insist on POS, require an out-the-door price disclosure before incentives, cap doc fees, and publicly report pass-through by dealer ID. Sunshine changes behavior very fast.
If I had to pick one mechanism to replace the current patchwork, here’s the simplest that hits outcomes and the used market:
- Utility-administered, pay-for-performance credit: Pay a flat, refundable $/kWh for verified EV charging, with a 2x multiplier off-peak and event-based bonuses for demand response/V2G. Fund it with a modest per-gallon carbon/road fee and/or a parallel feebate on high-CO2 registrations. Make it:
- VIN-tied and portable to second/third owners automatically.
- Measured via smart chargers, connected vehicle APIs, or approved telematics. For privacy, allow an odometer-based option with random audits.
- Capped annually per VIN (e.g., enough to cover 8,000-10,000 electric miles) with a higher cap for low-income households.
- Transparent: monthly credit on your utility bill or direct deposit if you can’t enroll the service address.
This directly rewards clean miles, steers charging to the right hours, and doesn’t care if the car was new, used, bought, or leased. It also scales with grid decarbonization automatically.
Guardrails to prevent gaming:
- One active credit per VIN, proof of residence for service address, and tamper checks on charger telemetry.
- Randomized odometer audits and cross-checks with utility interval data.
- Publish anonymized program stats so everyone can see kWh credited by zip code, VIN age, and time-of-use.
Feebate runner-up if you want a DMV lever instead:
- A revenue-neutral registration feebate based on vehicle test-cycle CO2 and expected lifetime VMT by segment, updated annually. You pay more for a Tahoe, get a rebate for a Bolt, smooth scale versus cliffy MSRP caps. Dealers can’t skim because it’s collected/paid at DMV.
On your data questions:
- Someone with access to J.D. Power PIN or similar could do the cleanest analysis on dealer capture. Look for price discontinuities at MSRP cap cliffs and before/after POS transfer start. My guess: capture exists, but varies widely by inventory tightness and brand finance policies.
- Lease vs buy: the only consistent “win” I’ve seen is when the captive heavily subvents the lease (inflated residuals plus buy-rate MF) and you’re certain you’ll exit before tires/brakes. If you’re paying marked-up MF and average fees, assume you’re getting 60-70% of that $7,500 in real value.
- Used-EV credits: without hold periods and VIN tracking, gaming happens. With them, you get decent targeting and fewer shenanigans, but programs still run out of money fast if they’re not capped or revenue-backed.
The tax code got us off the runway, but it’s clumsy at buying the outcomes we actually care about. Pay for kWh at the right times, modestly reward verified battery health down the road, and let DMV run a simple, transparent feebate. Fewer perverse incentives, better grid alignment, and the benefits follow the car into the used market where most drivers live.