Why are 5G electromagnetic radiation risks uninsurable?

Adrian Short · 2 March 2020

Some activists who believe that the the electromagnetic radiation from 5G mobile networks and similar technologies is unsafe make regular reference to the fact that insurance companies will not sell insurance to cover this risk. That itself is true. But these activists then go on to draw the conclusion that this is evidence that these technologies are unsafe. After all, their logic goes, if electromagnetic radiation from these networks were safe, why wouldn’t insurance companies cover them? And given that they won’t cover them, surely this means that the insurers think it’s not safe, or are sufficiently uncertain that they don’t want to take the chance. But this isn’t the conclusion that I draw and I think that these activists are misunderstanding how insurance works and getting the wrong end of the stick. So what makes a risk insurable or uninsurable?

Insuring cars against theft

Putting electromagnetic radiation to one side for a moment, let’s think about how insurance against car theft works. Some cars do get stolen and cars are generally quite expensive to replace, so it makes sense for most people to insure their cars against theft. An insurance company might sell car theft insurance to 50,000 people and charge every one of them a fee for that cover. If their car gets stolen, the insurance will pay for (most of) the loss. The insurance company knows that some of those 50,000 cars will get stolen but they don’t know which ones. And the insurer also has a good idea from previous experience broadly how many of those cars will get stolen and how much they’ll have to pay out for claims to insured drivers who’ve suffered a loss. Let’s say for the sake of argument that around 1% of those 50,000 cars will get stolen in any given year. So you’ve got 50,000 people paying their fees for insurance and the insurer will have to pay out to around 500 of them. If the insurer prices the insurance sensibly, they should make more than enough money from all 50,000 drivers to cover the expenses of running the business, compensate the drivers whose cars have been stolen, and make a profit on top. Of course, if the company prices the insurance too cheaply or if there are far more losses than expected, they might end up making a loss. If the insurance is too expensive, not enough people buy it to cover the costs of the business, the losses from theft, and still make a profit. That’s business. But insurance companies are experts in risk and the longer they stay in business the more data they’ve got to draw on to help to price their products appropriately and stay in profit.

Characteristics of poolable risks

Looking generally at this business, there are two key characteristics: There are expected to be far more “winners” - people whose cars don’t get stolen - than “losers” - people whose cars do. And the insurer generally expects that all or most of the losses are due to independent factors or events. Sure, there may be a few gangs of professional car thieves but generally speaking one person’s car getting stolen in London in March isn’t connected to someone else’s car getting stolen in Liverpool in September. One theft doesn’t cause or naturally follow from another. They are generally independent of each other. This approach is called risk pooling - a large number of drivers who are running a risk of their cars getting stolen are brought together by the insurer and given the opportunity to pay a relatively small fee that they can each afford to lose into a common fund so that any of them who suffer the major loss of their expensive cars can get compensated.

Now compare this to insuring against electromagnetic radiation risks from 5G networks. In principle, this radiation may or may not be harmful to human health. Certainly, many campaigners against 5G networks think that it is. And if they’re right, one of the consequences of installing 5G transmitters everywhere will be that very large numbers of people - potentially effectively whole populations - are made sick or killed by them. If that happened, the companies running this equipment would likely face millions of high-value claims for compensation, the total value of which would far exceed what they would be able to pay. Putting aside the very obvious catastrophic human cost in such a scenario, these financial losses would be catastrophic for those companies and put them out of business. This sounds like something those companies might want to insure against. And if the telecoms companies would want to buy insurance, why wouldn’t insurance companies want to sell it to them?

Now think back to the car insurance. The insurance company knows that most of the insured drivers won’t suffer a loss but a few of them will. How does that compare with a situation in which electromagnetic radiation from 5G networks is found to be harming or killing people? In that scenario, it wouldn’t be a few of the insured telecoms companies that were making claims to their insurers to cover their losses to people they had harmed, it would be all of them. And remember that in the car theft example, most of the car thefts have no connection to the other car thefts. One theft doesn’t generally cause or follow on from another. But if electromagnetic radiation is generally harmful, one claim for compensation from someone who had been harmed by it would be directly connected to the others: all the losses would all caused by the same thing - the intrinsic harmfulness of the technology itself.

Particular risk and fundamental risk

Insurers call things like the car theft example particular risk. The losses within the pool of insured risks are particular to the circumstances of every (or nearly every) individual car. This car gets stolen (for specific reasons), that one doesn’t (for different specific reasons). The electromagnetic radiation example is a case of fundamental risk. If the technology is fundamentally safe, no-one suffers a loss and the insurer doesn’t have to pay out to anyone. Also, all the people buying insurance have literally wasted their money. It would be equivalent to a world in which no cars ever got stolen or were even stealable. This wouldn’t be a guaranteed profit-maker for insurers, it would be an area in which they simply couldn’t sell insurance. But if 5G technology turns out to be fundamentally unsafe, all the telecoms companies that would have bought insurance would be facing massive losses all at once, and crucially, all for exactly the same reason. There’s no way that an insurance company, or quite probably the insurance industry, could cover losses on that scale. There’s no way to build a pool of companies taking a risk on electromagnetic radiation causing harm to human health where a few of those companies might cause harm (and therefore claim on the insurance) but most of them wouldn’t. It’s an all-or-nothing thing.

## Everyone’s in the same boat

I’m quite sure that there are some readers at this stage who feel that I have just proven that insurance companies are refusing to cover electromagnetic radiation risks because they suspect that these things really are dangerous and they don’t want to take the risk of paying out huge compensation. But that’s not my point at all. It’s not the size of the risk that concerns insurers nor even just the scale of the potential losses. It’s the shape of the potential pool of insured risks that makes electromagnetic radiation ininsurable. If these technologies are fundamentally safe then no-one needs to buy insurance and insurance companies can make no sales in that market. If the technologies are fundamentally unsafe, catastrophic losses are guaranteed and insurance companies can make no profits in that market. We already know this and the behaviour of the insurance companies doesn’t tell us anything new. Insurance companies don’t actually need to calculate the risk itself or take a view on how likely they think these technologies are to be unsafe. They are simply looking at it commercially and saying that the nature of the risk, however great or small it may be, isn’t something that can be profitably insured because a loss to one telecoms company due to electromagnetic radiation would mean a loss to all of them at once. Electromagnetic radiation liability among telecoms companies is a shared risk not a poolable risk, structurally similar but not necessarily of the same scale to two people in the same house facing the risk of burglary. It would be similar to a world in which there were a risk that all cars got stolen at once and for the same reasons, regardless of the size of that risk. That’s why they’re not selling insurance for it, not because they secretly believe that 5G networks aren’t safe.