Can Teachers Get Cheaper Car Insurance?

A Discount Is a Loss-Cost Statement, Not a Thank-You

Start with what a premium is made of, because the teacher discount lives inside one specific layer of it. Before expenses and profit load, a rate is built on the pure premium — the expected claim cost per exposure, calculated as claim frequency multiplied by claim severity. Frequency is how often a defined group files; severity is what the average filed claim costs to settle. Move either one down for a group and the group’s indicated rate moves down with it, automatically, before any discount is named.

So an occupational discount is not a marketing decision bolted onto a finished rate. It is a rate relativity: a filed statement that a class is cheaper to insure, which the insurer has to be able to defend. The professional rules make that explicit. Actuarial Standard of Practice No. 12, Risk Classification (for All Practice Areas)— adopted by the Actuarial Standards Board in December 2005, updated for deviation language effective May 1, 2011, and still the operative version — provides at section 3.2.1 that “the actuary should select risk characteristics that are related to expected outcomes,” and that “[a] relationship between a risk characteristic and an expected outcome, such as cost, is demonstrated if it can be shown that the variation in actual or reasonably anticipated experience correlates to the risk characteristic.”^[1]

The same section supplies the definition of fairness that the industry works under: “Rates within a risk classification system would be considered equitable if differences in rates reflect material differences in expected cost for risk characteristics.”^[1] Read that carefully. Equity here means the price gap matches the cost gap. It says nothing about whether the cost gap is one society wants to price on.

And here is the clause that generates the entire public fight over teacher discounts. Section 3.2.2 of the standard, titled Causality, provides that “[w]hile the actuary should select risk characteristics that are related to expected outcomes, it is not necessary for the actuary to establish a cause and effect relationship between the risk characteristic and expected outcome.”^[1] Nobody has to show that teaching makes you a better driver. The Actuarial Standards Board has proposed tightening that language: a September 2023 exposure draft of a revised ASOP No. 12 would move the discussion to a new section 3.2.3 and provide that the actuary “should have a rational explanation that the relationship between a risk characteristic and a risk measure is not obscure, irrelevant, or arbitrary; however, the actuary is not required to demonstrate a causal relationship.” That exposure draft has not been adopted, so the 2005 text governs.^[1]

What an insurer actually owes a regulator is set by state rate law rather than by the ASOP, and the two dominant regimes differ in when the state looks at the filing. California is a prior-approval state: Insurance Code § 1861.05(b) requires an insurer that wants to change a rate to file a complete rate application with the commissioner demonstrating that the change is justified, and § 1861.05(c) sets the notice and review clock that has to run before the application is deemed approved.^[21] Florida runs the file-and-use family: under § 627.062(2)(a), a filing made at least 90 days before the proposed effective date and not implemented during review is a “file and use” filing, while a filing made within 30 days after the effective date is a “use and file” filing that the office can later order refunded if the rate proves excessive.^[10] In both regimes the classification stands or falls on the support in the filing, not on a causal story about the class.

What the Behavioral Data Actually Measured

For decades the correlation between occupation and loss was a black box: insurers had claim outcomes by class and no visibility into the driving that produced them. Telematics closed that gap. Once a phone or a plug-in device is time-stamping braking events, an insurer can look at the behavior sitting between the rating variable and the claim.

The most detailed public version of that analysis is Behavioral Validation of Auto Insurance Rating Variables, co-authored in November 2021 by Dr. Robert Hartwig — Director of the Risk & Uncertainty Management Center at the University of South Carolina’s Darla Moore School of Business — and Robert Gordon of the American Property Casualty Insurance Association. It is a trade association publication written to defend these rating variables against restriction, and should be read as advocacy backed by data rather than as neutral regulatory research.^[2]

Its central observable is hard braking. The paper uses hard braking and hard acceleration as the behaviors linking driver characteristics to loss, on the stated basis that frequent hard braking and hard acceleration “have been proven to result in more frequent accidents.”^[2] What the paper then reports is an association rather than a mechanism: the groups that brake hard less often are the same groups whose claim costs run below the overall driver population, and the groups that brake hard more often run above it. It does not trace a causal path from any particular braking event to any particular collision, and it is written to defend these rating variables against restriction.

That is the frame in which the paper explains occupation, and the explanation is about routing rather than temperament: occupation “is predictive because it reflects vehicle usage and the regularity of commutes. For example, realtors generally drive more frequently and often to areas they aren’t as familiar with compared to a schoolteacher who tends to drive the same route to and from school every day. School teachers thus benefit from the use of occupation as a rating factor.”^[2]

Note what is and is not published there. The paper names schoolteachers explicitly as beneficiaries, and gives ranges for occupation as a class. It does not publish a teacher-specific percentage, and it cautions that “specific occupational comparisons may differ across insurers and states.” The example it does give from one insurer’s data is comparative: “Teachers benefit compared to financial advisors.”^[2] The paper also states that occupational risk groups “are not based on income. They are based on actual claim costs,” and that low-income and high-income occupations exist within each risk group — a claim consumer advocates dispute, as the next section covers.

The marriage row in that table is the one educators most often collect twice. If you are a married teacher, you sit inside two favorable relativities at the same time, which is part of why an educator quote can move so much more than a single advertised discount would suggest. Our separate research covers how marital status affects car insurance and, on the household side, whether good grades lower car insurance for a student driver on the same policy.

Why “Discrimination” Means Something Different on a Rate Filing

The sentence at the center of every rate filing is a model-law formulation rather than a national rule. The National Association of Insurance Commissioners publishes a Property and Casualty Model Rating Law (File and Use Version), GDL-1775, providing that “Rates shall not be excessive, inadequate or unfairly discriminatory.”^[3] A model binds nobody until a legislature enacts it, and states adopt it unevenly: Florida carries the same standard in its own enacted text at § 627.062(1).^[10] The operative word is unfairly, and the model supplies its own definition: “Unfair discrimination exists if, after allowing for practical limitations, price differentials fail to reflect equitably the differences in expected losses and expenses.”^[3]

Read that definition backward and the logic of the teacher discount appears. Charging two groups the same price when their expected losses differ is the violation. Sorting by expected cost is the requirement. Under this framework, refusing to give a lower-cost class a lower rate is closer to the prohibited conduct than granting it is.

The same model contains the provision group programs are built on: “A rate is not unfairly discriminatory if it is averaged broadly among persons insured under a group, franchise or blanket policy or a mass marketed plan,” where such a plan means insurance offered to employees of particular employers or to members of particular associations or organizations, or to persons grouped in other ways — “except groupings formed principally for the purpose of obtaining such insurance.”^[3] Two limits sit inside that sentence. It protects the averaging only in a state that enacted the language, and it protects it only where the plan is tied to an employer or an association that the insurance is sold through. A teachers’ union that exists to bargain contracts is the kind of pre-existing organization the definition describes. A category an insurer invents and labels “educators” to sort applicants is what the closing exception removes. Whether any particular union program is lawful still turns on the enacted rating law of the state it is sold in.

Consumer advocates argue that much of the industry has been operating on the wrong side of that line. In a February 2019 petition for rulemaking filed with the California Insurance Commissioner, Consumer Watchdog argued that examples of so-called affinity groups include “generic categories of occupations such as lawyers, doctors, engineers, business professionals, college graduates, homeowners, and other ‘groups’ fabricated by insurance companies solely for the purpose of discriminating against motorists to whom the insurance companies do not want to sell insurance.”^[6]

The petition attached quoted premiums rather than arguing in the abstract. On a basic limits policy for two drivers with the same driving safety record and everything else equal, it reported that Farmers charged a factory worker 14.5 percent more than either an accountant or a physician, $1,523 against $1,330; that Progressive charged a factory worker 6.3 percent more than an attorney or physician, $878 against $826; and that GEICO charged a factory worker 14.7 percent more than the same driver described as a corporate CEO, $977 against $852. On education alone, it reported Progressive charging an office manager with a high school diploma $878 against $826 for the same driver in the same occupation holding an undergraduate degree.^[6]

Those are quoted premiums gathered by an advocacy organization for a rulemaking petition, not a regulator’s audit. But they make the stakes concrete and symmetrical: the same mechanism that prices a teacher down prices a factory worker up, and a rating system is a closed book. The credit-scoring debate runs on identical logic, which our report on whether good credit lowers car insurance traces through the same fairness arguments.

Three States, Three Different Answers

Insurance is regulated state by state, so “can a teacher get an occupational discount” has no national answer. Three states show the full range of how the question is being resolved, and they are moving in opposite directions.

Colorado legislated against the input.Senate Bill 21-169, signed July 6, 2021, added C.R.S. § 10-3-1104.9, which defines an external consumer data and information source as a data or information source used by an insurer to supplement traditional underwriting or to establish lifestyle indicators — and then lists what counts: “credit scores, social media habits, locations, purchasing habits, home ownership, educational attainment, occupation, licensures, civil judgments, and court records.”^[8] Occupation and educational attainment are named in the statute itself. The act prohibits an insurer, pursuant to rules adopted by the commissioner, from using those sources or any algorithm or predictive model built on them in a way that unfairly discriminates on race, color, national or ethnic origin, religion, sex, sexual orientation, disability, gender identity, or gender expression.^[8]

Crucially, the statute does not ban occupational rating outright. It converts it into a compliance obligation. The rules must require each insurer to disclose the data sources it uses, explain how it uses them, establish and maintain a risk management framework reasonably designed to determine whether that use unfairly discriminates, assess the results and act to minimize the risk, and provide “an attestation by one or more officers that the insurer has implemented the risk management framework … appropriately on a continuous basis.”^[8] The Division of Insurance adopted an amended Regulation 10-1-1 extending that governance and risk management framework to private passenger automobile insurers, effective October 15, 2025, alongside life and health benefit plan insurers. The quantitative testing regulation — the rule that would prescribe how insurers actually measure disparate outcomes — remains a draft released for informal comment.^[9]

California never allowed the input, but allows the group.Under Insurance Code § 1861.02, enacted by Proposition 103 in 1988, auto rates are determined in decreasing order of importance by the insured’s driving safety record, annual miles driven, and years of driving experience, plus only “those other factors that the commissioner may adopt by regulation and that have a substantial relationship to the risk of loss.”^[4] Occupation is not among the optional factors the commissioner has adopted.

And yet educator programs operate in California, because a separate section of the same initiative opens a different door. Section 1861.12 provides in full: “Any insurer may issue any insurance coverage on a group plan, without restriction as to the purpose of the group, occupation or type of group. Group insurance rates shall not be considered to be unfairly discriminatory, if they are averaged broadly among persons insured under the group plan.”^[5] The Department has been working through what that permits. Its second workshop draft text of regulation on group plans under Section 1861.12 would add a group membership rating factor and require that an insurer “shall use a group membership rating factor and no other method, including separate base rates, to capture any rating differential for the group.”^[7] That text is a workshop draft, not adopted regulation. What § 1861.12 itself establishes is narrower and clearer than the draft: an insurer may issue coverage on a group plan without restriction as to the purpose of the group, occupation, or type of group, and group rates averaged broadly among the plan members are not unfairly discriminatory. What it does not do is add occupation to the § 1861.02 factors an insurer may apply to an individual applicant.

Florida leaves it to the actuarial support.Florida Statutes § 627.062 states that rates “may not be excessive, inadequate, or unfairly discriminatory,” and directs the regulator to consider past and prospective loss experience within and without the state, past and prospective expenses, trend factors including trends in actual losses per insured unit, conflagration and catastrophe hazards and projected hurricane losses where applicable, a reasonable margin for underwriting profit and contingencies, and other relevant factors affecting the frequency or severity of claims.^[10] There is no prohibited-input list of the Colorado kind and no mandatory-factor hierarchy of the California kind. A classification stands or falls on whether the filing supports it.

The Route That Survives Regulation Runs Through Your Membership Card

Trace the incentives on both sides and the union program stops looking like a perk catalog entry. The insurer wants concentrated access to a class it has already measured as low-cost, without having to defend an occupational rating variable in every state it writes in. The association wants a benefit its members can price and compare. In states that enacted the NAIC model language, a bona fide association is the kind of grouping the group-averaging provision covers, and California § 1861.12 permits the group plan in its own enacted text. Nobody has to argue about job titles, because eligibility is a verifiable fact.

The National Education Association runs the largest of these. Its NEA Auto & Home Insurance Program is offered in cooperation with Travelers, and NEA Member Benefits publishes the figure directly: “NEA members can save an average of $611* on auto insurance from Travelers.”^[17]

The American Federation of Teachers routes members differently. AFT’s own benefits page describes “access to special group discounts on auto and home insurance” and directs members to SnapQuote Insurance, which forwards to the Farmers Insurance Choice marketplace.^[18] Farmers GroupSelect publishes an average savings of $581 on that path, with a disclosure worth reading before you treat the number as a forecast: it is “based on average nationwide annual savings of new Farmers GroupSelect® customers surveyed from 1/1/24 to 8/31/25 who switched their auto policies to Farmers branded policies, responded to the survey, and realized savings.”^[19]

That disclosure is narrower than it looks, and Horace Mann’s $877 rests on a different population again: the company discloses it as an average annual savings amount based on new customers who provided their prior carrier’s policy details with effective dates from 1/1/2025 through 12/31/2025, which is a statement about whose numbers were compared, not a statement that those customers switched and realized savings.^[12] The Farmers figure is the one limited by its own disclosure to surveyed customers who switched and saved; drivers who got a worse quote and stayed put are not in that denominator. Either way the number tells you the program is worth a quote; it does not tell you what your quote will be. The only figure that means anything for your household is the one on a comparison quote built from your own declarations page, which is why our guidance on how often you should shop for car insurance and on whether car insurance can be negotiated applies with full force here.

Why a Whole Company Was Built Around This One Class

If the educator class were only marginally cheaper, no insurer would organize itself around it. One has, for eighty years, and its federal filings state the reasoning in plain language.

Horace Mann Educators Corporation began in Springfield, Illinois in 1945, “when two school teachers saw a need to provide automobile insurance to their association members” — note that even the founding was an association-based program.^[11] In its Form 10-K for fiscal year 2024, the company reports that the U.S. Department of Education estimates approximately 7.5 million K-12 school teachers, administrators and support staff nationwide, that Horace Mann serves approximately 1 million of these households, and that its customer base is about 80 percent educators, with the remainder generally in other public sector occupations such as firefighters.^[11]

Then comes the sentence that explains the discount from the insurer’s side: “The niche educator market has similar characteristics and preferred risk profiles, which allows for more precise underwriting processes and more targeted marketing operations.”^[11] Homogeneity is the asset, in the company’s own framing: what the 10-K claims the narrow class buys is more precise underwriting and more targeted marketing. The filing does not quantify a rate effect or say the carrier holds less margin for uncertainty, and neither does any other document cited here.

The auto policy is also not where the company makes its living. Horace Mann describes itself as a multiline financial services company with three reporting segments — Property & Casualty, Life & Retirement, and Supplemental & Group Benefits — and states that “Educators continue to begin their relationship with Horace Mann through 403(b) retirement savings products … which provide encouraging cross-sell opportunities.”^[11] Read the discount in that light. A competitively priced auto policy is the front door to a multi-decade relationship covering home, life, disability, and retirement annuities. That is not a reason to refuse the discount. It is a reason to price the other products separately rather than accepting the bundle on the strength of the auto quote.

The Part That Is Not a Discount at All

A standard personal auto policy insures the vehicle, not the property being carried inside it. When a teacher’s own laptop or class materials are stolen out of a parked car, the coverage that ordinarily responds is the personal property coverage on a homeowners or renters policy, which follows the insured’s belongings away from the home but carries its own deductible and its own claim record. Property the school owns is a separate question again, turning on the district’s own property coverage and on whether the teacher’s policy reaches property of others held off-premises. Read your own declarations page before assuming either answer.

Educator endorsements are built to close exactly that gap, and the published limits are specific. Liberty Mutual states that its teachers’ auto insurance covers teaching materials or school-owned property stolen or damaged while in the vehicle “up to $2,500 per occurrence,” applies a “$0 deductible” if the vehicle is vandalized on school property or during school-related events, and applies a “$0 deductible” if the vehicle is damaged in a collision while being driven on school business — with a footnote that a $100 deductible applies in Pennsylvania.^[13]

Horace Mann packages the same category of benefit as Educator Advantage. Its 10-K states that for both auto and property coverage the company offers educators “a discounted rate and the Educator Advantage® package of features,” including “liability coverage for transporting students in an insured vehicle and reimbursement for stolen school fundraising items.”^[11] Its consumer page adds coverage if work-related items are stolen or damaged while in the vehicle and waived deductibles for certain covered incidents on or near school property, describing all of it as available to “eligible policyholders” and subject to policy terms, conditions and limits, without publishing dollar figures.^[12]

Price these against the discount rather than on top of it. A $0 comprehensive deductible on a vandalism claim in a school lot is worth whatever your deductible is — commonly $500 or $1,000 — on the one occasion it applies, and nothing in the years it does not. A quote that is $60 cheaper but drops a $2,500 personal property grant is not obviously the better policy. Compare endorsement schedules, not just premiums.

Where a School Schedule Beats a Job Title

There is a second route to the same discount, and it sidesteps the occupational rating fight entirely. The white paper’s stated reason occupation predicts loss is vehicle usage and the regularity of commutes. A telematics program measures usage and driving behavior directly, for the individual driver, which reaches the same underlying facts without a regulator having to accept occupation as a rating variable at all.

Look at what the programs actually score. Progressive tells Snapshot participants to limit hard brakes and accelerations, avoid late-night driving — specifically between 12 and 4 a.m. on weekends — drive less overall, and stay off the phone.^[16] That is the whole of the mechanism: the program prices your recorded mileage and your recorded braking, accelerating, cornering, and phone handling, not a class average. No source cited on this page measures how educators as a group perform in these programs, and none of the three carriers publishes an occupational breakdown of its results.

The programs are not interchangeable, and the difference that matters is what happens when the data is unflattering. Nationwide frames SmartRide as one-directional: “The safer you drive, the higher the discounts you could get – up to 40%,” with an instant discount at sign-up, and states that SmartRide “is a discount program” that measures driving behavior “to reward you with a discount, not raise your rates,” while noting that in Virginia and Washington the discount may decrease at renewal based on changes in driving behavior.^[14] State Farm offers a 10 percent initial participation discount and savings up to 30 percent, and states that the adjusted amount “may increase or decrease at each renewal.”^[15] Progressive is the most direct about the risk: “Your rate may increase with high-risk driving. But, only about 2 out of 10 drivers actually get an increase.”^[16]

Enrolling also means handing a carrier a continuous record of where the vehicle goes and when. That is a separate decision from the discount, and it deserves its own consideration — our research on whether insurance companies can track your car sets out what these programs collect and how the data is used.

The One Trip That Can Undo All of It

Commuting to the building is ordinary personal use, and a personal auto policy is written for it. Loading two students into your back seat because the activity bus left and their ride never came is a different act in insurance terms — it is the employer’s work, performed in your car, and the question of whose policy answers has a specific and unwelcome sequence.

School districts protect themselves against this exposure with hired and non-owned auto liability coverage. Read what a district tells its own employees that coverage does. The Jersey Shore Area School District’s student transportation request form states that the district “will provide Hired and Non-Owned Automobiles Liability Insurance for the purpose of protecting the School District and its employees while they are operating personal vehicles for district business,” and then immediately qualifies it: “This insurance is secondary in nature, and is only triggered by legal accusations of negligence against the School District.”^[20]

Follow the order of operations that form describes. Under it, the employee’s personal policy answers first and the district’s policy sits behind it, conditional on the district itself being accused of negligence. If a claim exceeds the employee limits and the district is not named, there is nothing behind you. The same form makes the second gap explicit: “Insurance coverage for damages to the employees personal or rented vehicle is not provided by the School District, and should also be addressed by the employees Personal Auto Policy.” The district’s coverage pays other people’s injuries. It does not repair your car.^[20]

That district also sets a floor and assigns responsibility in writing. Employees wishing to transport students in personal or rented vehicles must provide proof of liability insurance of not less than $300,000 combined single limits, or $100,000 per person, $300,000 per accident, and $100,000 property damage in split limits, in the form of a current policy copy or a certificate of insurance — and by signing, the employee is “acknowledging primary responsibility for liability assumed by the transportation of students in a personally owned or rented vehicle.”^[20] Every statement above is that one district’s. Districts write these forms differently, and the only documents that establish where your own coverage sits are your district’s form and your own policy.

This is also the reason educator carriers write student transport into the base policy rather than leaving it to chance: Horace Mann’s 10-K names “liability coverage for transporting students in an insured vehicle” as an Educator Advantage feature.^[11] If you drive students even occasionally, that single line item may be worth more than the entire premium difference between two quotes. The general boundary between personal and business use is worked through in our research on when you need business car insurance.

Working the Question in Order

1. Start with the association, not the job title. In California, Insurance Code § 1861.12 permits group coverage without restriction as to the purpose of the group, occupation, or type of group; in states that enacted the NAIC model group-averaging provision, a bona fide employer or association group is the grouping that provision covers.^{[3] [5]}
2. Treat every advertised average as a reason to quote, not a forecast. Farmers GroupSelect discloses its $581 as the average among surveyed new customers who switched and realized savings; Horace Mann’s $877 is built from new customers who supplied prior-carrier policy details.^{[19] [12]}
3. Compare endorsement schedules alongside premiums.Liberty Mutual’s published $2,500 per-occurrence personal property grant and $0 school-property deductibles are coverage, not discount, and they do not appear in a premium comparison.^[13]
4. Check your own state before assuming the occupational discount exists. Colorado names occupation and educational attainment in its external consumer data statute; California limits auto rating to adopted factors; Florida leaves the question to the actuarial support behind the filing.^{[8] [4] [10]}
5. If the discount is unavailable, let the telematics program find the same behavior.The commute pattern that makes the occupational class cheap is directly measurable — but read the rate-increase terms first, because they differ sharply between programs.^{[14] [16]}
6. Resolve the student transport question before you need it.A district’s non-owned auto coverage can be secondary, conditional on the district being accused of negligence, and silent on damage to your own vehicle.^[20]