RESEARCH DOSSIER
Motor Insurance Demand Affordability and Uninsured Driving
Abstract. Motor liability insurance performs a public compensation function while being purchased, financed, and renewed by households facing unequal incomes, liquidity, risk exposure, information, and access to insurance. This monograph examines how premium burdens and market accessibility affect insurance demand and uninsured driving, how uninsured participation is measured, and how the resulting costs are distributed among drivers, insurers, victims, and public or collective guarantee mechanisms. It integrates classical insurance-demand economics with primary motor insurance legislation, official American affordability and uninsured-motorist evidence, California's low-cost insurance program, British claims and enforcement data, European Union motor insurance requirements, and Japanese compulsory bodily-injury protection. The analysis differentiates annual premium-to-income ratios from liquidity-constrained ability to pay; vehicle registration from actual road exposure; uninsured motorists from underinsured motorists; claimant compensation from underlying accident frequency; and associations from identified causal effects. The principal conclusion is that maintaining legally required coverage depends jointly on a sustainable premium relative to disposable resources, access to a suitable contract, credible compliance systems, and the effective protection delivered to accident victims. Targeted affordability interventions can improve accessible coverage, but voluntary application data alone cannot establish net reductions in uninsured driving. Administrative matching, exposure-sensitive measurement, distributional assessment, and credible causal evaluation are required to establish those effects. The dossier offers an original comparative analytical synthesis and an evaluation program, without claiming a new theory or an independently estimated treatment effect.
Keywords: Motor insurance; automobile liability insurance; insurance demand; affordability; premium-to-income ratio; uninsured motorists; underinsured motorists; financial responsibility; liquidity constraints; risk classification; road safety; victim compensation; public guarantee funds; insurance enforcement; California Low Cost Auto Insurance; comparative insurance regulation; causal inference.
Introduction
Motor insurance makes a privately financed contract responsible for a substantial part of the consequences of a socially shared activity. Drivers generate exposure on common roads; the physical and economic harms of a collision may fall on people who never chose the driver's insurance status. The insurance decision therefore combines household budgeting, expected loss allocation, accident law, enforcement, information, and the continued capacity of institutions to compensate victims.
The Global Status Report on Road Safety 2023 estimates 1.19 million road traffic deaths annually worldwide. That number measures mortality, not insurance penetration. It establishes the scale of road harm against which the compensation problem must be studied; extrapolating a global uninsured-driving count from it would be methodologically unsound. Exposure and insurance institutions vary substantially among countries.
Compulsory liability insurance changes the economic choice without eliminating it. A driver may face a statutory duty to insure and still confront a premium that cannot readily be financed. Noncompliance may also stem from mistaken assumptions about coverage, administrative interruptions, limited market access, deliberate evasion, or vehicles used outside the formal registration system. Understanding its prevalence requires observing the correct population and specifying what counts as uninsured.
In the United States, the Insurance Research Council's uninsured and underinsured motorists study reported an estimated 15.4 percent uninsured-motorist rate for 2023. This is an estimate based on insurance claims relationships, not a census of every person who drove uninsured. In the United Kingdom, the Financial Conduct Authority's Motor Insurance Claims Analysis describes a different system: compensation costs associated with uninsured drivers, collectively funded through industry arrangements. The indicators answer different questions.
Research consequently must distinguish three linked outcomes: whether eligible households can obtain and maintain the minimum legally adequate contract; whether uninsured vehicles and drivers actually participate in traffic; and whether injured third parties receive prompt and adequate compensation. No single one of these variables establishes the other two.
The inquiry proceeds through economic foundations, measurement, international institutions, distributional effects, and causal inference. Sources were selected from original statutes and regulatory documents, official administrative reports, published peer-reviewed economic studies, and identified industry research. Where the full text of a proprietary research report was inaccessible, only its publicly released findings and methodology were used. No individual-level proprietary insurance database was analyzed and no empirical causal coefficient is estimated in this monograph.
Scientific Definitions and Primary Foundations
The insured act, the insured person, and the insured vehicle
An insurance contract may attach to a vehicle, a named driver, specified permitted users, or a legally defined use of a vehicle. A compulsory insurance duty may apply to road use, public places, registered keepers, or particular classes of motorized vehicles. Accordingly, the term uninsured must always be interpreted relative to the operative legal rule, policy wording, time, use, driver, and relevant risk. A vehicle can appear in an administrative registry as insured while a particular use or driver falls outside the contract's conditions. An apparently uninsured registered vehicle may instead be lawfully withdrawn from road use.
The Road Traffic Act 1988, section 143 supplies the statutory basis for compulsory third-party motor coverage in Great Britain, including the requirement connected to use on a road or other public place under the revised provision. The government's continuous insurance enforcement guidance additionally distinguishes insurance requirements for registered keepers from cases where a vehicle is kept off the road under a statutory off-road notification. The legal statuses must not be conflated when measuring uninsured driving.
The European Union's consolidated Motor Insurance Directive establishes a common framework for third-party motor liability insurance and injured-party protection, while leaving significant contract and implementation details to national law. The amendments in Directive (EU) 2021/2118 address, among other matters, the definition of vehicles and use, compensation, and claims-related protections. A harmonized legal framework does not establish a uniform premium or uniform national enforcement intensity.
Japan illustrates a further distinction. The Ministry of Land, Infrastructure, Transport and Tourism's CALI system overview specifies compulsory automobile liability coverage for bodily injury and explicitly excludes property damage from that compulsory layer. Separate voluntary insurance can cover additional liability and property risks. Therefore, being covered by the mandatory scheme is not equivalent to possessing comprehensive motor insurance; a coverage-gap inquiry must identify the layer of protection at issue.
Voluntary demand inside a compulsory regime
Classical insurance economics explains why an individual may value risk transfer. Mossin's analysis of rational insurance purchasing provides an expected-utility foundation for demand under risk aversion, premium loading, and differing loss distributions. Ehrlich and Becker's market insurance and self-protection analysis further distinguishes purchased insurance from efforts to reduce the probability or severity of losses. These are general economic foundations: neither paper directly estimates the contemporary uninsured-driving rate.
A compulsory system adds legal consequences to the private risk-transfer choice. A driver who contemplates going uninsured faces possible fines, loss of vehicle use, recovery actions, civil liability, and the risk of catastrophic personal debt. These influences depend on the probability of detection, the expected consequences of a crash, the enforcement mechanism, and the capacity to pay the premium when due. The mere size of a nominal penalty does not reveal its expected behavioral effect.
A useful analytical decomposition is to distinguish (a) the driver's expected private value of risk transfer, (b) the expected cost of legal noncompliance, (c) the immediately payable premium and associated transaction charges, and (d) the driver's liquid resources and feasible alternatives. This is a decomposition of established economic considerations, not a proposed new demand theory. A person may value liability protection highly and yet be unable to purchase it at the relevant payment date. Another may afford coverage but choose deliberate noncompliance because perceived detection is weak.
Insurance demand and market participation are separate margins
Demand for a higher deductible, a lower liability limit, or optional collision insurance is different from the decision to retain the legally required minimum. Price changes may alter a policyholder's chosen coverage without changing whether the vehicle remains legally insured. Conversely, a high premium at the minimum legal limit may push an otherwise insured household out of the formal insurance system altogether.
The NAIC auto insurance overview describes underwriting, rating, coverage choices, deductibles, vehicle characteristics, mileage, and prior coverage as relevant components of the U.S. market. These factors change both the premium offered and the set of contracts accessible to an applicant. A study that observes only average premiums among insured policies will omit applicants who received unacceptable quotations, were denied, or never entered the market.
The distinction also matters for public policy. A program may reduce the mean premium paid by its participants through restricted coverage or eligibility and still have a limited effect on the wider uninsured-driving population. Research must establish how many new participants were previously uninsured, how long they maintain coverage, and whether increased insurance enrollment corresponds to actual on-road exposure.
The public compensation function
Liability insurance differs from purely first-party cover because a claimant's welfare can depend on somebody else's insurance decision. The economically relevant cost of uninsured driving is therefore broader than unpaid premiums. It can include incomplete judgments against at-fault drivers, uncompensated injuries, claims-handling costs, financing levies, litigation, and indirect burdens on other insurers or public systems.
The EU's motor victim-compensation provisions require arrangements addressing losses involving uninsured or unidentified vehicles, with national institutional implementation. Japan's government compensation guarantee program similarly addresses eligible victims of uninsured or hit-and-run vehicles. Such arrangements may protect victims while distributing part of the monetary cost of individual noncompliance across a wider population. The adequacy, eligibility, financing, and recovery rights of these mechanisms must be assessed separately.
Uninsured Driving Prevalence
The denominator determines what is being estimated
Uninsured driving is a state of vehicle use or driver conduct during a defined period. Several often-reported indicators measure different underlying populations. The proportion of registered vehicles without matched policies measures registry coverage; the proportion of licensed drivers uninsured in a survey measures a self-reported status; the proportion of vehicle-miles driven without required cover measures actual exposure; and the share of crashes involving uninsured motorists measures a crash-selected sample. None is interchangeable without assumptions about registration completeness, vehicle usage, enforcement, and crash risk.
The Insurance Research Council's U.S. uninsured-motorist estimates report 15.4 percent for 2023, with state rates ranging from 5.7 percent in Maine to 28.2 percent in Mississippi. The Insurance Information Institute's published methodological series describes the estimate using the ratio of uninsured-motorist claim frequency to bodily-injury liability claim frequency. Its table runs through 2022 and reports 11.1 percent for 2019 and 13.9 percent for 2020, whereas the IRC's later 2017–2023 research update reports 15.4 percent for 2023. The published values come from different study vintages and must not be concatenated into one unrevised time series. These figures reflect a claims-based estimation method and should be reported as estimates.
A claims ratio depends on policyholder populations, claim reporting, coverage take-up, accident circumstances, and claim handling. It is especially hazardous to treat the estimate as equivalent to a randomly sampled percentage of all person-miles driven. State comparisons remain informative only to the extent that coverage practices, exposure, and measurement conventions are sufficiently comparable. Published state rankings alone do not establish the effect of one state's fines or premiums.
Uninsured and underinsured risks must be separated
An uninsured motorist lacks the relevant liability cover at the time of the accident; an underinsured motorist possesses some cover that proves insufficient relative to a compensable loss or a jurisdiction's triggering definition. A vehicle may also be properly insured for statutory purposes while having only a modest liability limit. Underinsurance therefore depends on the size of losses, the selected limits, and the claims definition in use.
The IRC's underinsured-motorist findings estimated an 18.0 percent U.S. underinsured-motorist rate for 2023, distinct from its 15.4 percent uninsured estimate. The underlying concepts use different insurance and claim relationships. Adding them mechanically to infer a headcount would require evidence that the denominators, definitions and categories are mutually exclusive on the same observational basis.
For policy design, minimum insurance participation and adequate compensation are two distinct goals. Increasing minimum limits may improve expected victim compensation for insured losses but can raise the premium of minimum-compliant contracts. A credible assessment of any limit change must estimate both the additional protection and its incidence across policyholders, including potential coverage lapses.
Administrative verification and enforcement
Britain's vehicle enforcement policy documents regular matching between vehicle-registration records and the Motor Insurance Database, with advisory letters followed by enforcement measures in relevant cases. This is a concrete administrative mechanism: the identity of a registered keeper, coverage records and statutory off-road status are compared to identify potentially uninsured vehicles.
Registry matching can improve detection of sustained registration-related noncompliance. It still does not perfectly identify every uninsured road journey. Policy records can lag, vehicles may be used temporarily or across borders, and noncompliance by a particular unauthorized driver may not appear in a simple owner-policy match. Administrative matches should be evaluated for false matches, completeness, response delay, exemptions and changes in actual insurance purchasing.
The British government's guidance for uninsured vehicles establishes that a vehicle genuinely kept off the road under the proper declaration is treated differently. Consequently, using the stock of all registered but policy-unmatched vehicles as an uninsured-driving estimate would overstate the target without lawful-status and exposure adjustments.
Why cross-country comparisons require institutional normalization
The EU's compulsory third-party regime, Britain's keeper-based continuous verification, and Japan's compulsory bodily-injury scheme share a victim-protection goal but differ in insured perils, definitions, administration, and the nature of the statutory minimum. Japan's mandatory CALI enrollment rules make its insurance certificate and vehicle-type requirements a specific institutional arrangement, while optional cover may provide materially greater liability protection.
A comparison of compliance should therefore report at minimum the country and legal date, required insured hazards, coverage limits, types of vehicle included, relevant vehicle-use definition, exposure denominator, verification practice, sanctions, guarantee mechanism, and the data source's observation window. A lower nominal uninsured rate can coexist with a different population definition or with narrower mandated cover.
Identifying the causes of noncompliance
Observed noncompliance can be organized analytically into four mechanisms: inability to finance coverage, inability to obtain an acceptable compliant offer, informational or administrative failure, and deliberate risk-bearing under weak expected enforcement. These mechanisms may overlap. The classification is an organizing description of known economic and administrative issues, not a newly proposed theory.
Cross-sectional data cannot determine which mechanism dominates. A household's income may be correlated with risk location, car characteristics, household transport needs, driving record, and insurer rating factors. Higher measured uninsured prevalence in economically stressed areas is consistent with an affordability mechanism but does not identify its causal size. To establish mechanism-specific effects, research needs matched household, vehicle, premium-offer, policy, lapse, enforcement, and driving-exposure evidence, subject to lawful data governance.
Distributional and Safety Outcomes
The incidence of costs across households and victims
Insurance premiums are paid by policyholders, but the ultimate burdens of noncompliance can be dispersed across accident victims, public institutions and the insured population. If a victim cannot recover damages from an uninsured at-fault driver, the shortfall may be borne personally, financed through their own coverage, or addressed by a compensation institution. These alternatives redistribute losses differently even when the underlying crash frequency is identical.
The UK's Financial Conduct Authority's uninsured claims-cost figures report costs of claims associated with uninsured drivers rising from £329 million in 2019 to £452 million in 2024. Related Motor Insurers' Bureau levies rose from £331 million to £507 million over the period. The levy is a funding assessment; it is not analytically identical to claims paid in a particular calendar year. Cost changes can also reflect claims severity and processing, not merely the number of uninsured driving events.
In a January 2026 parliamentary answer, the UK Department for Transport reported a MIB-associated uninsured-driving claims cost corresponding to approximately 2.2 percent of the average premium paid by policyholders in 2024, based on the FCA's July 2025 analysis. The FCA document separately gives 2.78 percent as a figure associated with 2023 costs in its discussion. Year, denominator, and measurement definition must be preserved; neither percentage should be presented as a global levy rate or a precise causal premium increment.
A distributional evaluation must ask who funds the guarantee and who receives payment. The presence of a collective compensation mechanism improves resilience for eligible victims while shifting part of the cost toward policyholders or public finances. Any policy intended to lower uninsured participation should be evaluated for effects on uninsured victims, insured victims, low-income drivers who comply, and drivers who remain outside the formal market.
Victim protection under European and Japanese systems
The EU Motor Insurance Directive provides a legal foundation for compensation bodies addressing victims harmed by uninsured or unidentified vehicles. The 2021 amending directive addresses additional victim and insolvency protection, among other matters. Legal harmonization establishes duties and structures; empirical evaluation still needs national claim-payment data, processing times, eligibility rules, and the actual implementation record.
Japan's government guarantee for uninsured collisions provides compensation within the compulsory bodily-injury framework for eligible hit-and-run and uninsured motor vehicle victims. The compulsory insurance coverage description specifies bodily-injury categories and stated benefit ceilings. A comparison with British uninsured vehicle compensation must therefore normalize the coverable injury and loss types, not merely total amounts paid.
Formal eligibility does not guarantee a complete economic restoration after catastrophic injury. Tort recoveries, statutory ceilings, benefit coordination, the ability to pursue at-fault drivers, and claims access affect residual loss. The public protection objective includes both timely administration and the actual value of compensation after restrictions and claimant costs.
Poverty, mobility and exposure to road risks
A car may be essential to work or caregiving where transport alternatives are limited. A household confronting an expensive policy may trade off legally required coverage against employment-related mobility and other necessary expenditures. The distributional problem is therefore about the opportunity cost of compliance as well as the premium-to-income ratio. This mechanism is plausible and consistent with standard budget analysis; its prevalence requires household and transport evidence.
The Federal Insurance Office's geographic affordability analysis explicitly highlights traditionally underserved populations at the ZIP-code level. Its cross-sectional estimates can identify areas meriting further research but cannot assign a particular individual's motives or identify a racial, geographic or income characteristic as a causal determinant of illegal behavior.
From the insurer's perspective, refusal to differentiate expected claims cost may affect cross-subsidization and price adequacy. From the policyholder's perspective, heavy reliance on predictors associated with economic vulnerability can restrict access. The NAIC's account of insurance credit scoring and the actuarial risk-classification standard identify separate descriptive and professional aspects of this issue. The equitable policy question is how affordable minimum coverage and reliable compensation can be financed within the applicable legal and actuarial constraints.
Insurance status and actual road safety
The relationship between insurance and driving safety is not a simple substitution. Insurance may make compensation more reliable while weakening a driver's direct financial exposure to some accident losses. Enforcement may affect both insurance purchasing and driving behavior. Changes in claim reporting and the mix of covered drivers can alter observed accident rates independently of underlying road safety.
Cohen and Dehejia's state-panel study of traffic fatalities analyzed 1970–1998 U.S. data using an instrumental-variables strategy and reported evidence of moral-hazard effects of automobile insurance and no-fault liability rules. The study's historical setting, identification assumptions, and legal changes prevent treating its estimated effects as current universal coefficients. It nevertheless demonstrates why observed growth in insured participation cannot automatically be interpreted as a reduction in crashes.
The WHO road safety report provides the global public-health context, but it does not establish a quantified effect of insurance affordability on traffic deaths. An adequate safety study should distinguish collisions, injury severity, compensation, driving exposure, and reporting propensity. A decrease in uninsured claim frequency may reflect a different population of insured motorists or a change in claim reporting rather than safer driving.
Coverage expansion and the California evidence
The California program's reported increase from 46,175 active policies in 2024 to 66,998 in 2025 is a substantial increase in program participation, documented in the 2026 California report. The reported 77 percent of uninsured applicants is evidence about the selected applicant pool, not the share of the state's driving population newly insured. It also does not establish how long coverage persisted, whether the applicant would otherwise have purchased another policy, or whether total uninsured vehicle-miles fell.
In evaluating an income-targeted program, one must account for risk eligibility and selection. Applicants may be relatively safer drivers who meet qualifying criteria; other uninsured drivers may be ineligible or unaware of the program. Observed changes in aggregate statewide uninsured estimates during the same period may reflect macroeconomic conditions, premium changes, enforcement, migration, registration, and crash-claim reporting. Proper effectiveness analysis must isolate the program's incremental effect.
Evidence, Alternative Models, and Methodological Limits
A reproducible source hierarchy
Evidence for this dossier is organized by what it can establish. Statutes and official regulatory guidance establish operative requirements within a defined jurisdiction and date. Administrative statistics establish reported transactions or claim amounts under a specified counting rule. Academic economic studies identify particular theoretical implications or empirical effects under explicit assumptions. Industry estimates supply externally produced indicators, with the degree of independent verification limited by public access to the method and underlying data.
The law itself, rather than a secondary summary, is the proper primary source for the scope of compulsory cover. The EU consolidated directive, the British Road Traffic Act section 143, and Japan's official CALI explanation establish different statutory and institutional objects. Cross-country inference should begin only after their scope and enforcement differences are recorded.
The principal American aggregate uninsured estimate comes from an industry research program. The IRC public report summary was available, but the complete proprietary report and microdata were not independently replicated here. The transparent Insurance Information Institute time series corroborates the published series and explains the broad claims-frequency basis. That is sufficient to state the published estimate and its definition, not to claim independent verification of underlying claim records.
Formal measurement relations and sensitivity
Let U be the true proportion of on-road vehicle-miles driven without the required liability cover within a jurisdiction and period. Let R be an administrative proportion of registered vehicles without a policy match, and C a claims-based uninsured-motorist frequency indicator. U, R, and C differ because their populations and sampling processes differ. Equality would require strong assumptions about the probability of driving, the completeness of registration and insurance data, and the relative likelihood of claim-producing crashes or reported claims.
Suppose, illustratively, that uninsured and insured motorists have different accident propensities. Then a crash-selected uninsured share equals the uninsured exposure weighted by uninsured crash propensity, divided by the sum of uninsured and insured exposure weighted by their respective propensities. It equals the underlying exposure share only when those propensities are sufficiently similar and claims reporting does not create further differential selection. This is an algebraic illustration of selection bias, not an estimate of the relative dangerousness of uninsured drivers.
Similarly, the arithmetic difference in California's active low-cost policies, 66,998 minus 46,175, is 20,823 additional active policies at two year-end snapshots. The percent change, 20,823 divided by 46,175, is about 45.1 percent. The result is a verified accounting comparison, not an estimated reduction in uninsured motorists. Changes in applications, cancellations and renewals cannot be treated as independent additions to the year-end stock.
Competing explanations and identification
A correlation between premiums and uninsured driving may arise from a price effect, but it may also reflect common local risk factors, rising claim costs, regulatory differences, or selection. Drivers' prior accidents can raise their insurance premiums while simultaneously affecting claims outcomes. Regressing the uninsured rate directly on observed average premiums without correcting for these mechanisms risks reverse causality and omitted variables.
Credible causal designs may exploit eligibility thresholds for affordable programs, changes in minimum policy limits, insurance-verification rollout dates, administrative data matching, or localized premium shocks. Each proposal requires proof that comparison groups are genuinely comparable and that the intervention does not induce simultaneous changes in accident reporting, coverage definition, or exposure. Regression discontinuity around an income-eligibility cutoff requires the running variable to be observed and sufficiently resistant to manipulation. Difference-in-differences demands credible preintervention trends. An instrumental-variable design requires a defensible exclusion restriction, not merely a correlated policy variable.
Individual-level panels linking premium offers, household finances, policy effective dates, payment lapses, driver authorization, registration and observed road exposure would be particularly valuable. They also raise privacy, consent, governance and linkage errors. A program evaluation should specify the population whose coverage can change, the treatment actually received, and the counterfactual condition.
Interpretation of reported national findings
The U.S. sources use materially different time frames and quantities. The IRC's 2025 affordability brief analyzes actual expenditure through 2022 and forward projections; the Federal Insurance Office's January 2025 report uses 2022 minimum-limit premiums and ZIP-code income measures; the IRC's uninsured study provides a 2023 uninsured estimate. Their dates and denominators prohibit treating the three as contemporaneous observations of the same household.
The UK FCA's 2025 review uses insurer claims data and MIB figures to discuss cost pressures. Its estimates are about costs and reported claims, not the exact number of uninsured journeys. Britain's continuous enforcement system provides administrative detection but must still be evaluated against actual road-use exposure.
The EIOPA Consumer Trends Report 2024 concerns consumer experiences and access across insurance categories in EU markets. Its survey evidence can illuminate perceived value and access while being kept separate from the legally mandated third-party liability exposure that is the object of this dossier. Survey coverage for other motor products must not be relabeled as compulsory motor liability insurance penetration.
Evidence gaps and reproducibility limits
No original nationwide driver survey, claims-record audit, individual premium quote panel, or household expenditure microdataset was assembled for this monograph. The proprietary IRC uninsured report was not acquired in full. The public 2026 California report contains administrative aggregates rather than individual trajectories; the FCA review is a defined insurer sample and institutional cost analysis. Formal comparisons and policy interpretations presented here are authored syntheses of those materials.
These limitations are substantive but do not invalidate clearly scoped propositions about documented statutes, official counts, published estimates, or the logic of measurement. They do restrict causal claims and cross-jurisdictional comparisons. A reproducible future study would publish its unit definitions, insurer and policy coverage, time reference, legal versions, denominator choice, linkage rules, suppression criteria, preanalysis plan, model diagnostics, and sensitivity tests.
Comparative Scientific Synthesis and Research Conclusions
What the documented systems establish
The comparative material establishes three durable propositions. First, a compulsory liability requirement can coexist with meaningful gaps in access and compliance. The U.S. IRC uninsured estimates, the British uninsured claims-cost record, and California's low-cost program statistics document different dimensions of that issue. None alone establishes a uniform international prevalence measure.
Second, affordability involves the price of legally adequate protection and the ability to maintain it, rather than merely an insurer's average premium or a household's annual income. The Federal Insurance Office analysis and the IRC affordability comparison offer complementary measures that become more informative when liquidity, household composition and minimum-contract terms are incorporated.
Third, insurance compliance and victim protection are linked but nonidentical outcomes. The EU motor insurance framework and Japan's government compensation mechanism recognize the need to protect people harmed by uninsured vehicles. The actual protection delivered depends on eligibility, coverable losses, limits, processing and funding.
Implications for feasible public policy
A coherent policy response should address the specific margin responsible for noncompliance. Where an applicant has no suitable offer, assigned-risk and residual market facilities may matter more than stronger sanctions. Where temporary liquidity is the binding problem, payment timing and renewal continuity are central. Where noncompliance is deliberate, verified detection and proportionate enforcement are relevant. Where loss severity drives market-wide premiums, claims-cost efficiency and risk reduction may help affordability without shifting costs invisibly.
California's documented low-cost policy channel demonstrates the administrative feasibility of attracting eligible applicants, including people reporting uninsured status at application. The evidence supports continued analysis of this channel, while a causal claim of lower aggregate uninsured driving would require a separate comparison design and actual road-use measures.
The British registration and insurance matching process demonstrates an enforcement mechanism capable of identifying certain policy-record gaps. Its relevant policy tradeoffs include timely correction, legitimate exemptions, safeguards against erroneous matches, and incentives to obtain insurance rather than merely pay penalties.
Risk classification can support sustainable pricing while also altering the affordability distribution. The ASOP No. 12 guidance directs actuaries to evaluate expected outcomes, legal constraints, practical use and adverse selection. Intervention design should explicitly account for insurers' solvency and the risk that unsustainable prices reduce future availability.
A minimum evaluation specification
An evaluation of a motor insurance affordability intervention should define: the legal minimum policy that counts as compliant; the eligible vehicle and driver population; observed premium quotes and payment terms; household resources; preintervention and postintervention coverage days; actual on-road exposure; enforcement dates and sanctions; and claims consequences for insured and uninsured accidents. These are research design requirements, not a new theoretical framework.
Its primary outcomes should distinguish the proportion of eligible exposure under compliant liability protection, the length of coverage gaps, the number of newly sustained policyholders, and the share of eligible victims receiving compensation. Secondary outcomes may cover premiums, transaction fees, claim severity, road safety and distribution across income groups. Results should be compared under alternative definitions of registered vehicles, actual drivers, vehicle-miles and accident involvement.
Transparent counterfactual evaluation is essential. Program participation, published premium averages and a decline in claims ratios are useful measurements, but each can improve without the intervention causing a meaningful reduction in uninsured traffic exposure. Reliable causal conclusions require the same care in defining the untreated alternative as in counting enrolled customers.
Research questions that remain open
The following questions can be investigated with proper linked data. How sensitive is sustained minimum-liability coverage to the initial premium installment compared with the full annual premium? What fraction of uninsured road exposure is attributable to unsuccessful underwriting applications rather than deliberate evasion? Do low-cost insurance programs primarily insure formerly uninsured drivers or substitute for other affordable policies? How do minimum-limit increases change net victim protection once possible coverage lapses are included? How do verification systems affect actual uninsured vehicle-miles, as opposed to recorded administrative compliance? Which groups finance collective victim compensation, and how does the burden evolve with claim inflation?
Answering these questions requires new empirical projects. The present monograph establishes their conceptual and documentary basis without claiming to have answered them by statistical estimation.
Conclusion
Motor insurance affordability is a determinant of the practical accessibility of compulsory liability protection. Its measurement begins with a specified statutory contract, an actual payment schedule, and the financial resources of households expected to purchase it. Average premium-to-income indices offer useful comparisons, while liquidity, underwriting access, and coverage persistence determine whether those averages describe the experience of exposed drivers.
Uninsured driving prevalence is an empirical quantity tied to a population and observation mechanism. Claims-frequency estimates, registration-policy matches, applicant declarations, and insured-policy stocks measure different dimensions of the problem. The United States' published 2023 uninsured estimate, California's growth in active low-cost policies during 2025, and Britain's documented compensation costs demonstrate the policy relevance of the topic while requiring careful separation of their denominators and causal interpretations.
Public victim protection and road safety form distinct additional objectives. Collective compensation arrangements can reduce the losses borne by eligible victims even when individual noncompliance persists. Insurance coverage does not itself establish safer driving, and compensation quality requires evidence about effective benefits, timeliness and financing.
The defensible conclusion is that sustainable insurance participation depends on a combined institutional environment: affordable and continuously payable minimum coverage, accessible insurance supply, accurate information, lawful and credible verification, proportionate enforcement, and effective remedies for injured parties. The international evidence supports this integrated interpretation. It does not support one universal affordability threshold, one comparable global uninsured-driving rate, or a single policy instrument guaranteed to achieve compliance. Future identification studies must connect household constraints and insurance transactions to actual traffic exposure and victim outcomes.
References
Actuarial Standards Board. 2011. "Actuarial Standard of Practice No. 12: Risk Classification (for All Practice Areas)." Revised May 2011. https://www.actuarialstandardsboard.org/asops/risk-classification-practice-areas/
California Department of Insurance. 2024. "California's Low Cost Auto Insurance." Consumer alert, July 23, 2024. https://www.insurance.ca.gov/0400-news/0102-alerts/2024/California-s-Low-Cost-Auto-Insurance.cfm
California Department of Insurance. 2025. "2025 Report to the Legislature & Consumer Education and Outreach Plan" (covering 2024). California Low Cost Automobile Insurance Program. https://www.insurance.ca.gov/01-consumers/105-type/95-guides/01-auto/lca/upload/CLCA_2025_Legislative_Report.pdf
California Department of Insurance. 2026. "2026 Report to the Legislature & Consumer Education and Outreach Plan" (covering 2025). California Low Cost Automobile Insurance Program, March 15, 2026. https://www.insurance.ca.gov/01-consumers/105-type/95-guides/01-auto/lca/upload/CLCA_2026_Legislative_Report_ADA.pdf
California Department of Insurance. n.d. "Private Passenger Automobile Insurance Report." Accessed November 14, 2026. https://www.insurance.ca.gov/0400-news/0200-studies-reports/0400-auto/
Cohen, Alma, and Rajeev Dehejia. 2004. "The Effect of Automobile Insurance and Accident Liability Laws on Traffic Fatalities." Journal of Law and Economics 47 (2): 357–393. https://www.journals.uchicago.edu/doi/10.1086/422978
Ehrlich, Isaac, and Gary S. Becker. 1972. "Market Insurance, Self-Insurance, and Self-Protection." Journal of Political Economy 80 (4): 623–648. https://doi.org/10.1086/259916
European Insurance and Occupational Pensions Authority. 2025. "Consumer Trends Report 2024." January 15, 2025. https://www.eiopa.europa.eu/publications/consumer-trends-report-2024_en
European Parliament and Council of the European Union. 2009, as amended through December 23, 2023. "Directive 2009/103/EC Relating to Insurance against Civil Liability in Respect of the Use of Motor Vehicles." Consolidated version. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A02009L0103-20231223
European Parliament and Council of the European Union. 2021. "Directive (EU) 2021/2118 Amending Directive 2009/103/EC." Official Journal L 430, December 2, 2021, 1–23. https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX%3A32021L2118
Financial Conduct Authority. 2025. "Motor Insurance Claims Analysis: Multi-Firm Review." July 2025. https://www.fca.org.uk/publication/multi-firm-reviews/motor-insurance-claims-analysis-multi-firm-review-2025.pdf
Great Britain. 1988, as amended. "Road Traffic Act 1988, Section 143: Users of Motor Vehicles to Be Insured." https://www.legislation.gov.uk/ukpga/1988/52/section/143
Insurance Information Institute. n.d. "Facts + Statistics: Uninsured Motorists." Figures attributed to the Insurance Research Council, 2017–2023. Accessed November 14, 2026. https://www.iii.org/fact-statistic/facts-statistics-uninsured-motorists
Insurance Research Council. 2025. "Personal Auto Insurance Affordability: Countrywide Trends and State Comparisons." Research Brief, March 27, 2025. https://insurance-research.org/sites/default/files/2025-08/2025%20-%20Personal%20Auto%20Insurance%20Affordability%2C%20Countrywide%20Trends%20and%20State%20Comparisons.pdf
Insurance Research Council. 2025. "Uninsured and Underinsured Motorists: 2017–2023." Report summary, February 2025. https://insurance-research.org/node/130
Japan, Ministry of Land, Infrastructure, Transport and Tourism. n.d. "Overview of Compulsory Automobile Liability Insurance & Mutual Aid." Accessed November 14, 2026. https://www.mlit.go.jp/jidosha/jibaiseki/en/about/overview/
Japan, Ministry of Land, Infrastructure, Transport and Tourism. n.d. "How to Join Compulsory Automobile Liability Insurance & Mutual Aid." Accessed November 14, 2026. https://www.mlit.go.jp/jidosha/jibaiseki/en/about/policyholder/index.html
Japan, Ministry of Land, Infrastructure, Transport and Tourism. n.d. "Government's Program Guaranteeing Compensation for Automobile Accidents." Accessed November 14, 2026. https://www.mlit.go.jp/jidosha/jibaiseki/en/public_payment/index.html
Mossin, Jan. 1968. "Aspects of Rational Insurance Purchasing." Journal of Political Economy 76 (4, Part 1): 553–568. https://www.journals.uchicago.edu/doi/10.1086/259427
National Association of Insurance Commissioners. n.d. "Auto Insurance." Insurance Topics. Accessed November 14, 2026. https://content.naic.org/insurance-topics/auto-insurance
National Association of Insurance Commissioners. 2026. "Credit-Based Insurance Scores." Insurance Topics, updated March 19, 2026. https://content.naic.org/insurance-topics/credit-based-insurance-scores
UK Department for Transport. 2026. "Uninsured Driving: Written Parliamentary Answer 102305." Answered January 13, 2026. https://questions-statements.parliament.uk/written-questions/detail/2026-01-02/102305
UK Driver and Vehicle Licensing Agency. n.d. "Vehicle Enforcement Policy." GOV.UK. Accessed November 14, 2026. https://www.gov.uk/government/publications/vehicle-enforcement-policy/vehicle-enforcement-policy
UK Government. n.d. "Vehicle Insurance: Uninsured Vehicles." GOV.UK. Accessed November 14, 2026. https://www.gov.uk/vehicle-insurance/uninsured-vehicles
U.S. Department of the Treasury, Federal Insurance Office. 2025. "Report on Personal Auto Insurance Markets and Technological Change." January 2025. https://home.treasury.gov/system/files/311/Report%20on%20Personal%20Auto%20Insurance%20Markets%20and%20Technological%20Change.pdf
World Health Organization. 2023. "Global Status Report on Road Safety 2023." Geneva: World Health Organization. ISBN 978-92-4-008651-7. https://www.who.int/publications/i/item/9789240086517/