Auto Insurance CPI vs Used-Car and Repair Prices: Did Premiums Keep Climbing After Input Costs Cooled?
On a 2019 = 100 scale, July 2026 motor-vehicle insurance CPI sits at 149.9 while used cars print 132.1 — a 17.8-point wedge even as repair inflation moderates.
Loading interactive charts…
When used-vehicle prices exploded in 2021 and 2022, analysts treated motor-vehicle insurance CPI as the next shoe to drop. Replacement-cost inflation, longer repair times, and rising claim severity all feed actuarial models with a lag. The question in 2026 is whether that lag has finally closed — or whether premiums kept climbing even after the used cars and trucks and body-work price indexes cooled.
The Bureau of Labor Statistics publishes all three as separate CPI-U strata. We rebase CUUR0000SETE (motor-vehicle insurance), CUUR0000SETA02 (used cars and trucks), and CUUR0000SETD01 (motor-vehicle body work) to their 2019 calendar-year averages (= 100) and race them on the same chart. The answer is blunt: yes, insurance kept climbing after the input-cost shock faded.
As of July 2026, motor-vehicle insurance sits at 149.9 on that scale — up roughly 50% versus 2019. Used cars and trucks print 132.1, only 32% above the pre-pandemic base. Body work lands at 144.5. The insurance-minus-used-car wedge is 17.8 index points and has widened steadily since 2023 even as year-over-year used-car inflation turned negative.
The dashboard above opens seven panels: a triple-track index race, decoupling and repair-wedge ribbons, a used-car scatter, lag-correlation bars, epoch averages, and a year-over-year pulse. What follows is the narrative behind those lines.
What the three indexes actually measure
These are not interchangeable “car cost” gauges. Motor-vehicle insurance CPI tracks the price consumers pay for private-passenger auto policies — premiums, fees, and policy charges as sampled in the BLS expenditure survey. It is a cash outflow index, not a claims-cost index.
Used cars and trucks CPI measures transaction prices for pre-owned vehicles in the retail market. It spiked when semiconductor shortages choked new-car supply and dealers marked up trade-ins. It fell when inventory normalized and credit conditions tightened.
Motor-vehicle body work CPI isolates collision-repair labor and parts at body shops — a closer cousin to claim severity than the used-car sticker, but still only one piece of the loss triangle. Towing, rental cars, medical payouts, litigation, and fraud sit outside this line item.
Rebasing all three to 2019 = 100 puts them on a common ruler without pretending they share the same basket weights. A household’s budget share for insurance is not the same as its share for used vehicles or repair bills.
The 2021–22 used-car shock and the insurance lag
The used-car index blew past insurance during the reopening surge. By July 2022, used cars peaked near index 152.9 (roughly 53% above 2019) while insurance was still near 106. At that moment, insurers were digesting prior-year loss experience, court reopenings, and rising repair bills — but statutory rate-filing calendars meant many premium increases had not yet reached policy renewals.
Correlation analysis on monthly rebased indexes shows a 0.55 same-month Pearson coefficient between insurance and used cars across the 2019–2026 sample. Shift used cars six months ahead of insurance and the correlation rises to 0.63 — consistent with actuarial pass-through on a lagged basis, not a mechanical month-by-month link.
The epoch bars in the dashboard collapse the story into eras. During the 2021–22 spike, average insurance readings near 103 trailed used cars near 139. In the 2023–24 cool-down, used-car averages hovered near 133 while insurance averages pushed into the 137 range. By 2026 year-to-date, insurance averages near 154 against used-car averages near 128 — the widest sustained gap of the post-2019 period.
July 2026: premiums still climbing
The latest overlapping month is July 2026. Insurance CPI raw level prints 856.08 (1982–84 = 100 base) versus a 2019 average near 571 — the 149.9 rebased index. Used cars at 184.66 raw compare with a 2019 average near 140 for the 132.1 index reading.
Year-over-year momentum has diverged sharply. Insurance YoY remains positive in 2026 even as used-car YoY hovers in low single digits. The peak insurance YoY observation in this window is 22.6% in April 2024 — months after used-car YoY had already turned negative for stretches of 2023.
That pattern matters for household budgets. A driver who bought a used truck in 2021 at an inflated sticker may have seen trade-in values soften by 2024, but the renewal notice kept rising. The CPI strata capture population-average prices, not any one garage or carrier, yet the direction matches consumer complaints: premiums did not retreat when wholesale used-vehicle indexes did.
The decoupling ribbon — insurance minus used cars in percentage points — hit a sample peak near 31.6 pp in February 2026 before easing slightly to 17.8 pp by July. Even the easing month leaves insurance far above its 2019 pace while used cars sit only modestly higher.
Body-work repair costs: closer, but not identical
Collision repair inflation followed a different path than used-car stickers. Body-work CPI lacks a clean monthly run — BLS publication has gaps, including the October 2025 appropriations lapse that suppressed several CPI cells — so quarterly averages smooth the noise.
Where observations exist, body work climbed with insurance through 2022–24 but has not kept the same 2025–26 acceleration. On the July 2026 print, body work at 144.5 sits 5.4 index points below insurance. The repair wedge panel shows that gap widening after 2023: insurers repriced for severity and frequency factors beyond body-shop line items alone.
Labor shortages at collision centers, higher paint and parts costs, and more complex vehicle electronics all pushed repair bills up — but those forces moderated before insurance CPI peaked. Regulatory approval delays, reinsurance costs, and geographic catastrophe loads do not appear in the body-work stratum yet still land in premiums.
Why insurance does not track sticker prices month to month
Actuarial pricing works on experience periods and rate filings, not spot wholesale auctions. A carrier sets 2026 premiums using loss data from 2024–25, regulatory allowed margins, and forward reinsurance quotes. Used-car CPI can fall 10% year over year while filed rates still climb because:
- Severity trends can worsen even when parts inflation slowstotal-loss thresholds interact with used-car levels nonlinearly.
- Frequency (miles driven, accident counts) moved independently of vehicle prices post-pandemic.
- Uninsured motorist and personal injury social inflation do not appear in motor-vehicle goods indexes.
- State regulatory lag means approved increases hit renewals in waves, smoothing spikes into multi-year stair steps.
The scatter plot of monthly insurance versus used-car indexes shows positive slope but wide vertical dispersion after 2023 — months where used cars fell modestly while insurance jumped. That is the visual signature of decoupling, not measurement error.
Annual snapshot
| Year | Insurance idx | Used cars idx | Body work idx | Ins − used (pp) | Ins − body (pp) |
|---|---|---|---|---|---|
| 2019 | 100.0 | 100.0 | 100.0 | 0.0 | 0.0 |
| 2020 | 95.4 | 103.2 | 103.6 | −7.8 | −8.2 |
| 2021 | 99.0 | 130.7 | 109.4 | −31.7 | −10.4 |
| 2022 | 106.8 | 147.3 | 123.1 | −40.5 | −16.3 |
| 2023 | 125.4 | 136.8 | 131.2 | −11.5 | −5.2 |
| 2024 | 147.7 | 128.7 | 133.2 | 19.0 | 14.6 |
| 2025 | 156.4 | 132.1 | 136.9 | 24.3 | 19.5 |
| 2026* | 154.4 | 128.4 | 144.1 | 26.1 | 9.1 |
*2026 figures are year-to-date averages through July; 2025 includes an October CPI gap.
The table makes the handoff explicit. Insurance trailed used cars during the spike years (negative gap columns in 2021–22), then overtook them in 2024 and 2025 as used-car inflation cooled. Body work sits between the two peers — more correlated with insurance than used cars are, but still below premium growth in 2025–26.
Reading the data: caveats and limits
CPI is not your renewal quote. The insurance stratum is a national average across policy types, coverage limits, and discount programs. A teenage driver in a coastal metro can see double-digit increases while the national index rises high single digits.
Used-car CPI is not Kelley Blue Book. BLS samples a specific basket of vehicle types and ages. Trade-in values for popular trucks can move differently than the index.
Body-work gaps reflect publication holes. Missing months are excluded from quarterly averages; do not treat sparse months as zero inflation.
Correlation is not causation. The six-month lead correlation supports a lag story but does not identify regulatory, weather, or litigation shocks.
October 2025 CPI cells are missing due to the federal appropriations lapse; year-over-year calculations that cross that window should be read cautiously.
No internal link chain is needed to make the point: three public BLS series, one rebasing method, and a persistent 2024–26 wedge answer the headline question. Did insurance keep climbing after used-car and repair CPIs cooled? On the latest prints, yes — insurance remains near index 150 while used cars have retreated to the low 130s and body work has stabilized in the mid-140s. Until loss experience, reinsurance, and regulatory filings turn decisively, that gap is the baseline, not the exception.