Guide
An auto insurance premium is not a single number that arrives out of thin air. It comes from two connected steps. First, the insurer decides whether it wants your risk and on what terms (underwriting). Then, if it will insure you, the insurer applies its state-filed rating model to your specific information (rating). Different insurers use different rating factors and different weights, so the same driver, in the same car, at the same coverage, can honestly be quoted different prices by different insurers on the same day[1][2].
NAIC describes the pricing process as two steps rather than one. The distinction matters because different consumer questions live at different steps.
One consequence of the two-step model: consumer questions that feel like they belong to the price (why did this insurer decline me, why did I move to a different underwriting subsidiary at renewal) actually belong to the underwriting step. Questions about which rating factor moved the number belong to the rating step. Both are real, and both live in the state filings the insurer submits.
Method
This is an explanatory model, not a literal step-by-step formula that every insurer runs in this exact order. Insurers use their own rating models under state-specific rules; the sequence below is a useful way to see what enters the calculation and where.
1. The insurer collects driver and vehicle information.
Names, driver's license numbers, dates of birth, and driving history for each rated driver; year, make, model, VIN, and use profile for each vehicle; the garaging address; the coverages, limits, and deductibles you want; and prior insurance history. NAIC recommends providing the same information to each insurer so quotes are comparable[5].
2. Underwriting determines whether and how the risk fits the insurer.
Underwriting checks whether the risk fits the insurer's eligibility rules and, if so, which of its underwriting entities will write the policy[1]. Consumers who see "we cannot offer you a policy in this program but we can offer one in a different program" are usually reading an underwriting outcome, not a rating outcome.
3. The insurer applies its state-filed rating model.
Each insurer files its rating model with each state department of insurance. The model identifies the factors the insurer uses and how it combines them into a base rate that then gets adjusted for the specific applicant[3]. Because models are insurer-specific and state-specific, no single formula fits every quote in the country.
4. Driver and household factors enter the calculation.
Driving record and the records of others covered by the policy, typically over the last three years, are common inputs[2]. Triple-I adds age as an input (mature drivers vs teens and drivers under 25) and describes gender as a statistical factor in claims data[4]. Where state law allows, insurers may also use a credit-based insurance score[2][3] and, in some states, level of education as a rating factor[2]. Availability of each driver-side factor depends on your state.
5. Vehicle, location, mileage and usage factors enter.
Where you garage the vehicle matters; urban areas tend to have higher rates than suburban or rural areas because of higher observed vandalism, theft, and accident frequencies[2][4]. Triple-I also identifies annual mileage and vehicle use (commute vs pleasure) as inputs, with commute and business use generally pricing higher than occasional pleasure use[4]. The vehicle itself (make, model, trim, safety features, theft profile) is part of the base rate.
6. Coverage limits and deductibles change the priced policy.
Once the applicant-specific rate is calculated, the coverages selected and their limits and deductibles shape the priced policy. Higher liability limits or lower physical-damage deductibles generally raise the premium; the reverse generally lowers it. Coverage-specific mechanics live in our coverage guides, and the reasoning about which tiers fit lives in Guide #8.
7. Discounts, surcharges, and telematics adjustments are applied.
Insurers apply state-filed discount and surcharge factors (multi-policy, multi-vehicle, paid-in-full, autopay, good student, defensive-driving-course, vehicle-safety, and others). Usage-based insurance programs, where the driver enrolls in a telematics product, feed additional data (miles driven, speed, time of day, and other factors) into the calculation[2]. Program mechanics differ by carrier and state; see the frozen Progressive, State Farm, and Allstate reviews for how each carrier's program actually works. Some programs are discount-only; others can adjust a rate in either direction where state rules allow.
8. The final premium is produced for that insurer, profile, and state.
The final number is specific to that insurer, that driver profile, and that state. A different insurer working from identical inputs can produce a different final number without either quote being wrong. That is the point of comparing at identical inputs, which is documented in Guide #2.
Do not read this eight-step Playbook as the exact order every insurer runs internally. It is an explanatory model. Insurers apply their own state-filed rating models; the calculation order in the machine is not the point, the inputs and the state-specific rules are.
Even when a shopper provides the same information to two insurers, they typically get two different quotes. That is not an error. It is a direct consequence of how the industry is structured.
The identical-inputs comparison method (documented in Guide #2) is designed exactly for this: hold the driver, vehicle, coverages, and terms constant, then treat the price difference as the honest answer to which insurer will write this specific risk more cheaply today[5].
A useful way to hold the factor list in mind is to sort it by what it actually describes: the driver, the vehicle, the policy, the location and regulation, and the specific insurer's model.
Two insurers that use the same list of factors can still weight them differently in their filed rating models. That is why the same driver, in the same car, at the same coverage can honestly be quoted different prices by different insurers on the same day[2].
NAIC treats credit-based insurance scoring as one of many rating factors that insurers may use in personal auto pricing where state law allows[3]. It is a predictive score used by insurers, distinct from a lending credit score, and its use in insurance rating is regulated at the state level[3][6].
State rules on credit-based insurance scoring vary and can change. Several states restrict or ban the practice in personal auto rating. This page does not publish a state-by-state list because statutes and regulator actions change; the current rule for your state lives with your state insurance department. Our Guide #1 and cost guide both use the same framing.
Usage-based insurance (UBI) programs invite the driver to share driving-behavior data with the insurer. NAIC describes UBI as examining driving habits including miles driven, speed, time of day, and other factors to determine insurance costs[2]. Program mechanics differ across carriers and states; some programs are discount-only, and several major programs now allow the rate to adjust in either direction where state rules allow.
Rather than duplicate carrier program details here, the frozen Provider Reviews document how each individual program actually works, including the state carve-outs and the two-way rate mechanics where they apply.
The important point at the rate-calculation level: a telematics program is one more input into the insurer's state-filed rating model, and enrolling in one does not guarantee a discount[2]. Read the program terms in your state before assuming what the program will do to your renewal.
The short answer: insurers file different rating models, weight the same factors differently, use different underwriting rules, refresh data on different schedules, and file different discount menus. Each of those is a reason for two insurers to produce two different final numbers for the same driver on the same day[2][3].
None of that makes the quotes wrong. It makes the identical-inputs comparison necessary. NAIC recommends providing the same information to each insurer at quote time so the difference between the numbers is a real difference between the insurers, not an artifact of differently-worded inputs[5]. The mechanics of that comparison live on Guide #2.
Renewal notices can show a different premium even when the driver, the vehicle, and the coverage all look the same. Two mechanics account for most of what shows up.
Alongside those two, the specific policy can pick up small changes without a driver noticing: an added or removed discount, a shift in the insurer's rating of a vehicle model, or a change in a rating factor the insurer refreshes on its own schedule (driving-record data, credit-based insurance scoring information where allowed, and so on)[2][3].
Renewal notices are required to identify the specific reason for a change in most cases; if the reason is not clear, ask the insurer in writing. If the answer does not satisfy, NAIC's Consumer Information Source is where consumers file complaints or check an insurer's record before shopping[7].
This Guide is explanatory. Once you know the mechanics, three practical moves cover most of what a consumer can do at the rate-calculation level.
Ready to see how another insurer prices the same driver profile at identical inputs?
Compare car insurance quotesNo. NAIC states directly that premiums for auto insurance are tailored to each individual and each insurer uses different rating factors[2]. Two insurers can use overlapping factor lists but weight them differently in their filed models.
No. Insurers file their own rating models with each state department of insurance[3]. There is no single national formula that produces every premium; the same insurer can price differently in two states because it is running two different filed models.
There is no single answer that fits every insurer, every state, and every driver. Different insurers weight the same factors differently in their filed models[2]; different states allow different factors[3]. Ask what moves your specific quote by comparing at identical inputs across insurers.
No. State rules on credit-based insurance scoring vary and can change. Several states restrict or ban the practice in personal auto rating[3]. This page does not publish a state-by-state list because the current rule for your state lives with your state insurance department.
No. Program mechanics differ across carriers and states[2]. Some programs are discount-only; several major programs now allow the rate to adjust in either direction where state rules allow. See the frozen Progressive, State Farm, and Allstate reviews for how each carrier's program actually works.
The most common answer is that the underlying rating model moved. Insurers file rate changes with each state department of insurance, and approved rate changes flow into renewal calculations regardless of what happened on a specific policy[3]. Book-of-business claim experience can drive those changes. The renewal notice is the place the insurer identifies the specific reason; if the answer is not informative, ask in writing.
Because each insurer files its own rating model, uses different underwriting rules, refreshes data on its own schedule, and files its own discount menu. NAIC: premiums for auto insurance are tailored to each individual and each insurer uses different rating factors[2]. That is why NAIC recommends providing the same information to each insurer so the price difference is a real difference between the insurers[5].
Your state department of insurance is the source of truth for which rating factors are allowed in your state and for what. Rules on credit-based insurance scoring[3], gender[4], education as a rating factor[2], and other factors vary and can change. This page does not publish state lists because they change; the state department of insurance publishes the current rule.
Regulator publication describing the underwriting-then-rating structure that produces a personal auto premium and the identical-inputs discipline consumers should follow when requesting quotes.
Regulator-topic overview of common personal auto rating factors: driving record over roughly the last three years, credit-based insurance score where allowed, garaging location, education where state law allows, and usage-based insurance that examines miles driven, speed, and time of day.
Regulator overview of credit-based insurance scoring as one of many rating factors in personal auto pricing where state law allows. Anchors the state-variation framing on this page; the current rule for a given state lives with the state department of insurance.
Consumer-education guidance on the driver-side factors insurers commonly use: driving record, annual mileage and vehicle use, driver age, driver gender (with statistical framing rather than a universal rating rule), and garaging location.
Regulator worksheet for consumers requesting side-by-side quotes; anchors the identical-inputs discipline that lets consumers compare insurer pricing on the same facts.
Regulator-authored consumer piece framing credit-based insurance scoring in plain language, including that state rules on its use vary.
Where to look up an insurer's market-share-adjusted complaint index by state. Useful when an unexplained renewal change makes a comparison at another insurer worth requesting.
The underwriting-and-rating structure, the identical- inputs discipline, and the common rating-factor categories are drawn from NAIC consumer guidance and NAIC insurance-topic pages. Driver-side factor descriptions (record, mileage, age, gender, location) are drawn from Triple-I. Credit-based insurance scoring is framed as one of several factors whose use is state-specific; this page does not publish a state-by-state list because statutes and regulator actions change. Gender and education are named as factors where state law allows, without a state list for the same reason. Telematics program specifics live on the frozen Provider Review pages. This page publishes no factor weighting percentages, no "most important factor" claim, and no universal formula. Last reviewed .