One of the classic questions in economics is about when the state should be involved in the market, and when not. In general, governments need to get involved when there are various types of market failures, positive or negative externalities, or natural monopolies. Some types of insurance require no government intervention and are left best to competing insurance companies. For other types of insurance there is a strong case for government involvement—either as a regulator (mandating service levels, comprehensive access, and other minimum standards) or as direct provider. The case for government involvement in the insurance sector hinges on whether there is a public good involved, such as population health (for health insurance) or road safety (for automobile insurance). The public insurer can have a broader mandate that includes prevention and treatment, not just covering financial losses.

In Canada, provincial government entities (crown corporations) operate comprehensive auto insurance in British Columbia, Saskatchewan and Manitoba, while Quebec's SAAQ only covers bodily injury and death. Other provinces have private insurance systems. The conventional argument in favour of private insurance is that public insurance companies do not innovate sufficiently because they don't need to, and that private insurers offer more choice at lower costs due to competition. Yet, the evidence is far from convincing that private automobile insurers keep costs low. It is simply not the case that the provinces with private insurers have lower insurance premia compared to the provinces with public insurers. Ontario and Alberta with private insurers have higher premia than Manitoba, Saskatchewan and Quebec with public insurers (Car Insurance Rates Across Canada: Who's Paying the Most and Why?). BC stands out with higher costs, but for reasons that have nothing to do with public provision—my August 2017 blog How to fix car insurance in British Columbia explained the main reasons why costs have gone up. However, insurance premia are also on the rise now in Alberta, Nova Scotia, New Brunswick, provinces where insurance is private. Meanwhile, in Manitoba (with public provision), car insurance rate are set to decrease by 0.6% in 2020.

The cost of all automobile insurance, private and public, is ultimately determined by the frequency and cost of claims. This means that premia reflect the cost of risk in the long run, and also the process through which claims are settled. When claims cost rise, private insurers simply balance the books and increase premia. The role of a public insurer is rather different, because the state has an interest in keeping the cost of mobility low and reduce the number of accidents to keep its citizens safe and sound. This points to a classic market failure: private insurers have no incentive to reduce accidents and improve road safety because they can simply pass on the cost of higher claims to drivers. They pool the risk but don't reduce the risk. Public automobile insurance internalizes the externality: the public insurer has a government mandate to keep roads safe, and is equipped with the tools to do so ranging from driver licensing, insurance policy features to encourage better driving, to deploying road safety equipment and helping with enforcement measures. Of course, road safety is a shared responsibility with other government branches, but the public insurer focused on cost reduction has a unique set of tools to so that is not available to private insurer. It is the very fact that it is a state monopoly that provides the advantage, because in a competitive market place there would be free-riding among insurers. If one company improved road safety in one spot, all other would benefit equally. A public insurer's objective is to minimize cost, not to maximize profits. Put another way, the ideal public insurer is one that shrinks in size over time because accidents become rarer and roads become safer.

Consider a simplified example. There is a particularly accident-prone intersection that leads to injuries and damages worth one million dollars a year. The private insurer covers the risk, and if that risk is spread across a million drivers, each driver pays $1 for this risky intersection through higher premia. Revenues match costs. The public insurer, however, has a different calculation. Fixing this intersection and spending an amortized $100,000 per year reduces the crash rate and costs drop to $400,000 per year, an assumed 60% reduction in risk. Total costs have thus dropped to $500,000 per year, and the public insurer can pass on the savings as a 50-cent reduction in premia to its customers. The difference between the private and public insurer is that the public insurer can also act on its mandate to improve road safety, based on benefit-cost calculations for each risk factor and location. Economic theory tells us that the insurer should fund any road safety initiative that has a positive net social benefit (i.e., present-discounted benefits exceed present-discounted costs). The insurer's additional role as road safety agent allows it to influence and reduce the risk. Society is better off as a result. To fully enable the potential of a public insurer, it is thus important to enable the public insurer with a strong road safety mandate and equip it with the tools to fulfill this mandate. This in turn requires rigorous cost-benefit analysis to target the most useful road safety interventions and reward the most effective vehicle safety features.

Private insurers also have a pernicious tendency to sort customers based on exclusion—in some jurisdictions they are permitted to turn down applicants. Insurers try to avoid adverse selection by motorists, who of course know more about their own driving skills and risk potential than the insurer. Thus private insurers try to observe consumer attributes to predict risk, and the more data they can use the better they can discriminate. In some jurisdictions, some attributes such as gender and age are legally excluded from being used. However, risks are segmented based on factors predictive of the probability of an accident. As these are predictive factors only, the unlucky applicant who is a safe driver but has the wrong characteristics may find it difficult to obtain affordable insurance. Even where governments mandate a take-all-comers (TAC) requirement for private insurers, exclusion happens effectively through stratification of premia. In some cases, an insurer-of-last-resort offers coverage for the residual market. By contrast, public insurers have a mandate to take all customers right from the start, arguably providing greater fairness with respect to accessing insurance. The risk pool is the entire population. Through exclusion, private insurers who are armed with better information than their competitors can create more profitable risk pools. Of course, some drivers are better off with private insurance, but at the cost of less fairness with respect to accessing automobile insurance affordably.

Arguments in favour of public insurance go beyond fixing the market failure that detaches road safety from insurance. A public insurer also benefits from larger scale and reduced transaction costs because multiple services are all under a single roof. A single public insurer also has a data advantage, because information on the cost of repairs and treatments from individual vendors can be compared against each other, limiting the potential for abuse and fraud. More importantly, a public insurer also has a mandate for accountability and transparency. The British Columbia Utilities Commission (BCUC) is tasked with regulating ICBC and approving basic insurance rate changes.

Public insurance is not unique to the automobile sector. Workers' compensation boards operate on the same principle, where the objective is to reduce accidents and improve worker safety, rather than simply to match injury costs with premia on an actuarial basis. The success of public insurers in this domain comes from the unique ability to pair insurance with accident prevention. The same principle should apply to automobile insurance.

Saskatchewan's Government Insurance (SGI) is a good example of how costs can be controlled. Created in 1945 as a crown corporation, SGI operates the Saskatchewan Auto Fund as the province's compulsory auto insurer, alongside driver licensing and vehicle registration. SGI has a mandate to operate on a break-even basis and does not receive money from the government or pay dividends to the government. Saskatchewan's Auto Fund has been able to control costs effectively by focusing on the treatment of injuries and getting people back to work, and aligning the goals of the insurer and injured person by linking benefits to commitment to the treatment plan. Avoiding costly adversarial settlements and replacing them with generous but outcome-focused treatment plans ultimately serves motorists better.

The main concerns about public insurers are related to transfers between the insurer and the government—subsidies to cover shortfalls, or dividends siphoned off as general revenue—and issues related to innovation. The first concern can be fixed by mandating a long-term break-even rule for the crown corporation that prevents positive or negative government transfers. The second concern is about a potential lack of innovation. There are two types of innovation, one that is important and one that is not. The "good" type of innovation designs insurance products that incentive good driving behaviour. The government mandate to keep roads safe also provides a key performance indicator (KPI) for the public insurer. Management should be rewarded for making our roads safer, thus aligning the incentive to innovate with the public objective. The less useful type of innovation differentiates the insurance product by offering "bells and whistles" that allow customers to customize their coverage based on risk aversity, but ultimately does nothing to make our roads safer. Public insurers are good at the first type of innovation, and private insurers are more focused on the second type of innovation.

In practice, public and private insurance can co-exist. In the health care sector we have universal basic insurance through the government and optional private extended insurance (for example, for extended benefits, dental care, or international travel). Similarly, extended coverage can be provided by private insurers of automobiles. In fact, many private insurers offer extended coverage in British Columbia, where consumers have a choice of using private insurers for comprehensive and add-on insurance in addition to the mandatory ICBC Autoplan coverage. Public and private insurance can complement each other.

In light of the high costs of auto insurance in British Columbia, some have argued that privatizing automobile insurance provides a fix. For proponents or privatization, increased competition is a panacea. However, the focus on privatization is a red herring. As pointed out above, the source of the high costs in BC has nothing to do with public or private provision; it is a result of escalating claims cost. The gains from competition advertised by a recent MNP report (see below) mostly arrive from bundling discounts, loyalty programs, and hypothetical innovation. The only significant suggestion is to improve the actuarial basis of rate setting through using age and actual amount of driving. Age as tool of discrimination is disallowed in BC, and thus this has again nothing to do with public versus private provision but with the regulatory framework. The only meaningful suggestion is to use actual mileage for setting premia—something perhaps worth considering, but which can easily be provided by a public insurer as well. Lastly, minor injury caps suggested in the report have already been implemented by BC's government, and again has nothing to do with public versus private provision.

The bottom line is that BC is actually very fortunate to have public automobile insurance. The task ahead is to strengthen public car insurance in BC, reign in the costs, and ultimately focus on improved road safety. Fewer accidents means lower costs and fewer injuries and deaths. Perhaps ICBC also needs to make its public mandate for road safety clearer in public discussions, and perhaps the provincial government should seek to strengthen and expand this road safety mandate. Society will be served best when roads are safe—for motorists, bicyclists, and pedestrians alike—and insurance costs for motorists are low as a result. The best road accident is one that never happens, and the best insurance is one that's never needed.