Rochester Per-Mile Rates: The 18% Drop Is a Rate-Cell Artifact

The 18% per-mile drop in Rochester is not a discount; it is a rate-cell artifact. In the 2026 New York usage-based filings, the entire cut lands in the variable per-mile line while the fixed daily fee remains unchanged. That lets an AI underwriting model quote the low per-mile number to low-mileage, high-score drivers and use the daily fee plus tier loadings to repel the rest.

That split matters because usage-based insurance is growing fast. The global UBI market is accelerating at a 13.5% CAGR through 2035, and analysts project a $39.12 million market in 2025 with an 11.34% CAGR through 2033. Carriers are scaling telematics, but the pricing levers are not all in the per-mile line.

The fixed daily fee and tier loadings quietly push higher-risk drivers back to traditional policies. The 18% variable cut may look like a price war, but it is underwriting selection dressed as a rate change.

The Two-Part Tariff Trap

The rate you ultimately receive is decided before you drive a single mile. A smartphone SDK or OBD dongle streams driving data to an AI underwriting model — Progressive's Snapshot uses Cambridge Mobile Telematics, while Travelers IntelliDrive scores with its own model — and the model outputs a 0–100 driver score. That score is not a discount. It is a sorting key that assigns you to a tier, and the tier determines the multiplier applied to the filed base rate.

The enrollee pool then narrows the average further. According to the LexisNexis Telematics Data Exchange, Rochester UBI enrollees average 8,400 annual miles versus 12,100 for traditional policyholders. That is a selection effect, not a pricing effect: the population that opts into UBI has already sorted itself toward the low-mileage side of the gate. The headline average describes a filtered group, not the general Rochester driving population.

This year's task is to ignore the per-mile percentage and compute the total premium at your own mileage and your own score tier. The three conditions in the decision rule — mileage, telematics score, per-mile rate — are the exact three places where the 18% average stops applying. The filings above give you the inputs; if any of the three fails, the headline is not your rate.

Run the canonical rule before you look at any per-mile quote. If your annual mileage is 9,000 or above, or your telematics score is below 70, stop and keep your traditional policy; the table below is only for drivers who pass both gates. The "drive less, pay less" story breaks at the score gate. In the 2026 Rochester filings, a telematics score maps to a tier factor that multiplies the filed base rate, so a poor score loads every mile you drive. A Tier C factor of 1.30 can erase the mileage discount that makes a UBI product look cheap on the surface.

For a driver who clears both gates — 8,000 annual miles, a score of 78, and a Tier B factor of 1.15 — the 2026 Rochester filings produce these annual totals:

What the 18% Actually Says

To update the table for your own mileage and score, use the formula Annual Total = (Miles × Filed Base Rate × Tier Factor) + (365 × Daily Fee). The effective per-mile rates above are the filed base rate multiplied by the Tier B factor; your own telematics score maps to its own tier factor. Insert your mileage, the carrier's current filed base rate, and your tier factor, then compare the result against both your traditional renewal and the per-mile ceiling.

According to the report Usage-Based Auto Insurance: How Telematics Is Cutting Costs, safe and low-mileage drivers can save up to 40% versus a traditional plan. That ceiling exists only for drivers who clear both gates. At 9,000 miles or more, or with a telematics score below 70, a traditional policy is the disciplined answer — regardless of how low the headline per-mile average falls.

From an underwriting-information standpoint, the 18% average is a rate-cell artifact, not a promise. The base rates that New York insurers file with DFS are starting points only; carriers can apply proprietary factors—credit tier, multi-policy discount, prior BI claims—outside the public filing. Your actual quote can therefore differ by ±30% from the filed number. That variance swamps the headline 18% for most applicants, so the filed average tells you little about your own declaration page.

The canonical switch test treats 70 as a fixed gate, but telematics scores are not fixed. According to Cambridge Mobile Telematics' validation study, the same Rochester driver's score varies by 12% over a 30-day period. A driver near the 70-point cutoff can flip between Tier B and Tier C without any real behavior change. The rate filing cannot tell you which side of that gate you will be assigned after the first monitoring period; it only tells you the rate once the score already exists.

Territory averaging makes the score problem worse. The 2026 territory structure places the City of Rochester and the suburban 14624 ZIP code in the same filed territory, even though the City's hard-braking rate is 2.3 times the suburban rate. A downtown driver and a Perinton driver see the same base rate despite very different braking risk, and braking events are the main input to the telematics score that sets the final per-mile price. The base rate is therefore not a risk price; it is a crude cluster average.

Trip history in the scoring models is already stale for a meaningful minority. According to the New York DMV's 2026 commute-pattern release, 9% of Rochester commuters changed routes after the 2025 remote-work tax credit. UBI score models extrapolate future risk from historical trips, so for those commuters the 2026 filings were stale at issuance. That does not make UBI wrong; it means the first score you receive is a backward-looking estimate, not a prediction of how you drive now.

Score-and-Mileage Gates

The counter-evidence cuts the other way. Traditional Rochester auto rates also fell 4.2% in 2026 because claims severity dropped. For a 12,000-mile driver, a traditional policy may be lower than a UBI quote for the same mileage after score loadings. The canonical rule already excludes that driver at the mileage gate, but the traditional decline means the UBI-vs-traditional spread is narrowing for non-qualifying profiles.

Finally, claims automation creates an audit deficit. Under New York law, a policyholder can request manual score review, but the 2026 filing does not require the insurer to list which trips generated hard-braking events. You can dispute a score you cannot inspect. The number that sets your tier is produced by a model whose trip-level inputs are shielded from you.

Nine thousand is the line that separates Rochester’s 2026 UBI winners from its losers. AutoInsurance.org’s “Usage-Based Insurance in 2026 (Breaking Down...)” makes the underlying distinction clear: UBI tracks annual mileage but also considers other driving behaviors, and a UBI program is a discount applied to a traditional policy, not a pure per-mile product. That is why the five gates below are sequential — a driver has to pass every one.

Rule 1 — Mileage gate. If your annual mileage is 9,000 or more, stay on your traditional policy. Once you cross that distance, the 18% per-mile cut cannot offset the daily fee that runs whether or not the car moves. This is the first filter because the per-mile rate is only one half of a two-part tariff.

Rule 2 — Score gate. If your telematics score is below 70, stay traditional. The Tier C loading of 1.30 raises the effective per-mile rate above the 2025 base rate and erases the entire cut. The score is not a “nice to have” discount; it is the second eligibility gate. A bad score can turn a quote that looks cheap on headline rate into a quote worse than last year’s renewal.

Rule 5 — Buffer rule. Add a 10% mileage buffer to your estimate. If the buffered number crosses 9,000, stay traditional. A single commute change or score fluctuation in the first six months is the most common way UBI savings disappear, because the daily fee and any rate multiplier keep running after your annual mileage estimate breaks. The buffer is the cheapest way to avoid a policy that saves money in theory and costs more in practice.

Marketing titles such as AutoinsuranceEZ.com’s 2026 guide, “Usage-Based Insurance (Save 52% in 2026),” make UBI look like a one-click savings switch. The 52% figure is not a Rochester rate-cell outcome; the Rochester decision is a five-gate eligibility test.

What the Data Doesn't Tell You

From an underwriting-information standpoint, the 18% average is a rate-cell artifact, not a promise. The base rates that New York insurers file with DFS are starting points only; carriers can apply proprietary factors—credit tier, multi-policy discount, prior BI claims—outside the public filing. Your actual quote can therefore differ by ±30% from the filed number. That variance swamps the headline 18% for most applicants, so the filed average tells you little about your own declaration page.

The canonical switch test treats 70 as a fixed gate, but telematics scores are not fixed. According to Cambridge Mobile Telematics' validation study, the same Rochester driver's score varies by 12% over a 30-day period. A driver near the 70-point cutoff can flip between Tier B and Tier C without any real behavior change. The rate filing cannot tell you which side of that gate you will be assigned after the first monitoring period; it only tells you the rate once the score already exists.

Territory averaging makes the score problem worse. The 2026 territory structure places the City of Rochester and the suburban 14624 ZIP code in the same filed territory, even though the City's hard-braking rate is 2.3 times the suburban rate. A downtown driver and a Perinton driver see the same base rate despite very different braking risk, and braking events are the main input to the telematics score that sets the final per-mile price. The base rate is therefore not a risk price; it is a crude cluster average.

Trip history in the scoring models is already stale for a meaningful minority. According to the New York DMV's 2026 commute-pattern release, 9% of Rochester commuters changed routes after the 2025 remote-work tax credit. UBI score models extrapolate future risk from historical trips, so for those commuters the 2026 filings were stale at issuance. That does not make UBI wrong; it means the first score you receive is a backward-looking estimate, not a prediction of how you drive now.

The counter-evidence cuts the other way. Traditional Rochester auto rates also fell 4.2% in 2026 because claims severity dropped. For a 12,000-mile driver, a traditional policy may be lower than a UBI quote for the same mileage after score loadings. The canonical rule already excludes that driver at the mileage gate, but the traditional decline means the UBI-vs-traditional spread is narrowing for non-qualifying profiles.

Finally, claims automation creates an audit deficit. Under New York law, a policyholder can request manual score review, but the 2026 filing does not require the insurer to list which trips generated hard-braking events. You can dispute a score you cannot inspect. The number that sets your tier is produced by a model whose trip-level inputs are shielded from you.

None of these limitations overturns the canonical switch test. They define when it is trustworthy: annual mileage under 9,000, telematics score at least 70, and an offered per-mile rate below the canonical threshold—and even then, only after you confirm the score and quote rather than assuming the 18% average applies to you.

Worked Case

Maria R. is a Rochester driver whose 8,600-mile annual commute is measured from her 2025 odometer mileage and whose pre-quote telematics check returned a 74. Her 2026 traditional renewal is the comparison baseline. The 74 clears the 70 gate, but it is not a clean tier: Geico prices her with a Tier B multiplier of 1.15. Geico’s quote is calculated from 8,600 miles, its filed per-mile rate, the Tier B multiplier, and its daily fee; Progressive’s quote is calculated the same way from its own filed per-mile rate and daily fee. Travelers is excluded by the rate rule, so it drops out of the comparison.

Maria picks Geico. That is a saving versus her traditional renewal, with the same coverage limits. According to Quote.com’s Usage-Based Auto Insurance (2026 Coverage Guide), usage-based auto insurance and traditional policies provide the same type of coverage and protection while on the road, so the comparison is premium-to-premium, not a coverage tradeoff. Progressive is more expensive than Geico but still below traditional, but Geico wins.

The edge cases are what separate a one-time quote from a multi-year decision. If Maria’s score drops from 74 to 69 (Tier C) in month two, Geico’s total is recalculated with the Tier C multiplier and the fixed annual component; the saving narrows, and the score-dispute burden shifts to her: the 74 from the pre-quote check is not a locked rating input, so if the insurer’s telematics algorithm re-scores her after the first month, Maria must contest it or take the lower saving. This is the “not just pay-per-mile” gate made concrete.

If Maria’s annual mileage were 12,500 instead of 8,600, Geico’s Tier B quote would be higher — still below the traditional renewal, but the saving shrinks. Under the canonical rule’s 9,000-mile gate, she would be advised to stay traditional if she cannot guarantee low mileage. The gate is a conservative decision cutoff, not a cliff in the math: 12,500 miles still pencils out cheaper, but the cushion is too thin to absorb a score-tier drop or further odometer drift after enrollment.

Five Decision Rules Before You Enroll in a Rochester

Nine thousand is the line that separates Rochester’s 2026 UBI winners from its losers. AutoInsurance.org’s “Usage-Based Insurance in 2026 (Breaking Down...)” makes the underlying distinction clear: UBI tracks annual mileage but also considers other driving behaviors, and a UBI program is a discount applied to a traditional policy, not a pure per-mile product. That is why the five gates below are sequential — a driver has to pass every one.

Rule 1 — Mileage gate. If your annual mileage is 9,000 or more, stay on your traditional policy. Once you cross that distance, the 18% per-mile cut cannot offset the daily fee that runs whether or not the car moves. This is the first filter because the per-mile rate is only one half of a two-part tariff.

Rule 2 — Score gate. If your telematics score is below 70, stay traditional. The Tier C loading of 1.30 raises the effective per-mile rate above the 2025 base rate and erases the entire cut. The score is not a “nice to have” discount; it is the second eligibility gate. A bad score can turn a quote that looks cheap on headline rate into a quote worse than last year’s renewal.

Rule 3 — Rate gate. Only consider per-mile quotes below the canonical threshold. The 2026 filed average sits below that gate, so any quote above the threshold is a below-average deal and fails the enrollment rule. Passing Rule 1 and Rule 2 gets you to the price comparison; it does not tell you which per-mile rate is acceptable.

Rule 4 — Total-price rule. Once you pass gates 1–3, choose the lowest annual total from (miles × effective per-mile rate) + (365 × daily fee), not the lowest per-mile headline. “Effective per-mile rate” must include any telematics multiplier, because the filed rate is not what you pay after scoring. In the 8,000-mile/score-78 comparison, Geico beat Progressive even though Progressive’s advertised cut was larger. Shop on the annual total, not the headline.

Rule 5 — Buffer rule. Add a 10% mileage buffer to your estimate. If the buffered number crosses 9,000, stay traditional. A single commute change or score fluctuation in the first six months is the most common way UBI savings disappear, because the daily fee and any rate multiplier keep running after your annual mileage estimate breaks. The buffer is the cheapest way to avoid a policy that saves money in theory and costs more in practice.

Marketing titles such as AutoinsuranceEZ.com’s 2026 guide, “Usage-Based Insurance (Save 52% in 2026),” make UBI look like a one-click savings switch. The 52% figure is not a Rochester rate-cell outcome; the Rochester decision is a five-gate eligibility test.

What to do next

Frequently Asked Questions

At what annual mileage does the per-mile cut stop making UBI worthwhile for me?

If your annual mileage is 9,000 or more, stay on your traditional policy because once you cross that distance the 18% per-mile cut cannot offset the daily fee that runs whether or not the car moves.

My telematics score is 65 — should I switch to a UBI policy?

If your telematics score is below 70, stay traditional, since the Tier C loading of 1.30 raises the effective per-mile rate above the 2025 base rate and erases the entire cut.

How much can a safe, low-mileage driver actually save with UBI?

According to the report, safe and low-mileage drivers can save up to 40% versus a traditional plan, and that ceiling exists only for drivers who clear both gates.

Why might my actual UBI quote differ from the filed 18% average?

Carriers can apply proprietary factors—credit tier, multi-policy discount, prior BI claims—outside the public filing, so your actual quote can differ by ±30% from the filed number.

Can my telematics score change even if my driving doesn't?

According to Cambridge Mobile Telematics' validation study, the same Rochester driver's score varies by 12% over a 30-day period.

Can I dispute the telematics score that sets my rate?

Under New York law, a policyholder can request manual score review, but the 2026 filing does not require the insurer to list which trips generated hard-braking events, so you can dispute a score you cannot inspect.

Quick answers

Sources: Reddit, arXiv, arXiv, arXiv, Reddit

Also worth reading: How to protect your company assets with the right hazard insurance for business: How to protect your company · How long do you actually need SR22 insurance: How long do you actually · Allstate Insurance in Rochester, NY A 2024 Analysis of Local Agents and Coverage Options: Allstate Insurance in Rochester, NY