Breaking News! Superman is fighting an unidentified person in downtown Metropolis!
Didn’t this happen last week? And isn’t this all getting a bit expensive?
Most of the time when property is damaged, the property owner has insurance that will pay to restore their property to approximately the state it was in before the loss occurred. But when Doomsday goes on a rampage of destruction across at least three states or the Joker blows up half of downtown Gotham, insurer’s aren’t actually going to want to pay for that, and there is reason to believe that under the terms of standard insurance contracts, they wouldn’t have to. The reason has to do with the way insurance policies are written, which is a matter of contract as much at it is a matter of law.
So the focus of this post is not whether supervillains can get insurance, but whether standard insurance policies will pay for damage that they cause.
I. Insurance Policies and the DICE Method
First of all, insurance policies are only written for insurable risks. Generally speaking, an insurable risk is one where both the probability and magnitude of a particular kind of loss are measurable, where the occurrence of that loss is truly random, and where it is possible to transfer that risk to an insurer for an economically-feasible premium. A common example of an insurable risk is one’s house burning down. We know how often houses in a particular zip code burn down (this is what actuaries do for a living; some fun, huh?), we know what a particular house is worth, houses don’t burn down at any predictable frequency, and as it turns out, it’s possible to insure against the risk of fire for a premium which is both acceptable to the insured and profitable for the insurer. Flood is an example of an uninsurable risk. Floods do occur at random, and we know basically how often, but the magnitude of losses caused by flood are such that it is impossible to offer flood insurance at any price a homeowner can afford (more on this particular exposure later). Floods are considered “catastrophic” losses, because they cause both a high amount of damage to individual properties but also a high amount of damage to entire regions, making it impossible to adequately spread the cost to other property owners. The same is true of war, terrorism, civil unrest, revolution, etc., which is why all of those are considered uninsurable risks. Discharge of nuclear weapons, intentional or accidental, is also uninsurable. Uninsurable risks are generally excluded from insurance policies.*
When a loss occurs, the claims adjuster is going to walk through the DICE method: Declarations, Insuring agreement, Conditions, Exclusions. First, look to see if there is coverage for this kind of loss on the declarations page, i.e. coverage scheduled for this particular policy. Then, check the insuring agreement to see if the loss results from a covered peril. Then, check to see if there are any relevant conditions in the policy which are applicable. Finally, see if there are any relevant exclusions.
Take the Doomsday example again, and let’s assume that he has just leveled a private residence insured by ABC P&C by throwing Superman through it. ABC’s adjuster is first going to look at the declarations page for the insured’s homeowner’s policy. The house is insured for $100,000. So far so good. Then, he’s going to look at the insuring agreement to see if there is anything of interest there. This policy is a special perils form, which covers everything not specifically excluded, so again, so far so good. Then he’ll check conditions. The homeowner is current on his premium, gave timely notice of the claim, and is cooperating with the adjuster, so again, probably okay there.
But what’s this? Terrorism is excluded? And you didn’t buy the terrorism endorsement? Hmm. That’s going to be a problem.
It’s going to be pretty easy to argue that Doomsday is a terrorist, but even if he isn’t, it isn’t going to be difficult to fit this into either the war or civil unrest exclusions, both of which are part of every insurance policy. Any insurance defense attorney worth his salt would certainly make that argument, and it’s hard to see why it wouldn’t win. Heck, if Superman is a state actor, it might be excluded under the “civil authority” exclusion.
So sorry, Mr. Homeowner, your insurance policy isn’t going to pay for this.
II. Uninsurable Risks and Residual Pools
So what’s to be done? If we’re talking about a universe where superheros and supervillains exist and unstoppable monsters do level significant sections of town every other Tuesday, it seems probable that the legal system and/or insurance industry would take this into account. But because the magnitude of losses caused by superhero battles are so great, it seems likely that the states would have to resort to residual market mechanisms. This is how flood insurance is currently offered on a national level: the FEMA National Flood Insurance Program NFIP is pretty much the only way to buy flood insurance anymore. States have set up residual markets for both high risk drivers and properties with significant windstorm (Mississippi, South Carolina, Texas, etc.) and earthquake exposures (California) too. Basically, state legislatures have decided that even though certain kinds of risk are impossible to insure against on the open market, we want people to take those risks anyway, for a host of possible reasons. We want high-risk drivers to be insured both for their protection and for others, and denying someone permission to drive because they cannot buy mandatory insurance seems unjust. People really want to live in earthquake– and hurricane-prone areas, and those people vote, so we’re going to find some way of making that work, no matter how silly it is.
Residual markets can work in one of a number of ways. One is “assigned risk,” an approach frequently used to ensure that high risk drivers have access to at least the state minimum liability limits for personal auto insurance. Basically, every insurer that participates in the market is required to take their fair share of high risk drivers–for a high premium–as a cost of doing business in the state. They can then spread this cost to their other insureds, keeping the companies profitable. But it seems more likely that the states would create their own “Supervillain Pool” similar to the windstorm pools active in Gulf and coastal states. The way these work is that every insurance policy is charged a tax based on the premium which goes into the residual pool. The pool then reimburses property owners for damages caused by supervillain rampages, etc. Property owners would need to buy “Superhero/Supervillain Insurance” from the pool, and that premium would help too, but because this truly is an uninsurable risk, the pool will probably need to be supported by taxpayer revenue. The idea is that all but the biggest losses will be at least mostly absorbed by the pool but that the government will step in if things get really out of hand. The pool can theoretically up its rates in the years following a big loss to ensure that the government gets its money back, but this rarely happens.
Of course, while we’re modifying the law to account for superheros, it would probably also be the case that insurance companies would include some kind of “superhero/supernatural/paranormal” exclusion, shifting that exposure more directly to the residual pool, as has been done with flood, earthquake, and windstorm exposures.
A world with recognized and regular supervillain rampages would probably develop a way to insure against that sort of thing, but traditional property insurance would probably exclude such losses. States would need to create residual pools, much like the way they have for earthquake and hurricane losses.
*”Speculative risks,” i.e. risks where there is a possibility of gain and a possibility of loss–investments, basically–are also uninsurable, because the cost of insuring them would more than erase the potential gain from the transaction. These aren’t excluded by insurance policies because they wouldn’t have been covered anyway; insurers simply don’t write policies for those kinds of risks.
Commentary by James Daily:
I agree with my co-author that comic book worlds have likely developed a way to insure against attacks by supervillains and collateral damage from the response by superheroes. However, even if the cost is evenly distributed among the risk pool, the question of net social cost remains. In other words, are superpowers a net social good or a net social cost? I think that, on balance, superpowers are either neutral or slightly positive.
Empirically, it seems that most versions of the DC and Marvel universes are broadly similar to our own in terms of the standard of living and technology level. If superpowers were a net social cost we would expect one or both to be lower relative to the real world. Indeed, if anything the technology level is sometimes higher in comic books, although advanced technology is often confined to superhero and supervillain labs. Whatever the reason (superheroes fighting regular crime, inventions from Batman, Tony Stark, and Reed Richards), superpowers do not seem to be a net social cost. This fact supports the civil rights claims of superpowered individuals.