I'm curious how this math works. In my (quite likely incomplete) mental model of insurance, there are only a few terms: premiums, payouts, profits. It would seem that they have to balance. So if you pool risk, you are summing the payouts. Those payouts are covered by the sum of premiums. Any profit is the sum of premiums minus the sum of payouts.
Profits = Premiums - Payouts
Is this right so far?
Because then it seems that the expected value of insurance to the customer is the negated value of the expected profit of the insurer. Put another way: if the expected value of insurance is not negative, shouldn't the insurer raise premiums or exit the business?
Again, I know next to nothing about insurance and I'm asking to learn.
The point of insurance is to smooth risk, not to pool it. Pooling risk is not the point of insurance as a product; it's the mechanism by which that product (risk smoothing) becomes feasible as a business model.
> Look up the term "actuarially fair".
There's really no need to condescend by telling me to look up basic terminology; I am quite well-versed in this topic.
I didn't bring up moral hazard in this post, because it actually isn't necessary to explain why insurance has a negative expected value (which is equivalent to the statement that premiums are always non-zero[0]). It sounds like you're confusing a few different concepts here.
[0] Moral hazard is generally addressed by deductibles, not by premiums.
On an infinite timescale it may not add up but it surely can on any practical one, right?
from an individual perspective, insurance means you don't have to worry about unexpected expenses from things beyond your control (ie, peace of mind).
Now, the mechanism by which this is achieved may be described as a pool in that it's a single company with a pool of money that everyone pays into and the input into that pool necessarily must be greater than the output.
But from the "customers" perspective, insurance is peace of mind. ie, it smooths their total risk over time.
You will always have to worry about unexpected expenses, because insurance companies are never going to insure everything that could possibly happen, even if you get a good policy that doesn't dick you around on claims.
If you are expecting insurance to protect you from not knowing the risks, then you are taking a much bigger risk than not having insurance at all, because of the complacency that the belief you're insured engenders.
What I'm looking for is the ELI5 for what insurance really is about (especially from the POV of insurers) given that, if I understand correctly, 'chimeracoder is disputing the insurance-as-everyone-pooling-money model.
people and business purchase insurance to cover unexpected loss, ie, "peace of mind".
Maybe insurance companies are evil scumbags, but that's STILL the reason why people PURCHASE insurance.
But everyone seems to forget that there's a huge part of that pool taken by the things sitting in the insurance office's chairs.
And no, I'm not kidding.
There is a broad market in specialist insurance which caters to the non-volume cases. It just costs more because you have to expend more effort-per-sale (and of course because non-volume also tends to mean that the risks genuinely are higher - e.g. homeowners with prior subsidence claims).
Of course if no one else is looking why give up the easy money - after all it is not like the customer can buy a new policy without the massive markup. Collusion does not need to be explicit, it can be accomplished just through everyone using the same systems and rules.
Competition only works if at least one of the parties involved is trying to compete. If all the insurance companies are happy to slap on a blanket risk premium then there is no competition. This is great if you want to disrupt the status quo.
For car insurance it makes sense, as the risk is specific to the driver, so the number of claims tells you something about the insured's driving.
But, for home insurance, isn't it mainly about the property rather than the individual? And isn't that mostly about location and property type?
Will this continue to affect her premiums if she moves? Will it also affect the new owner of her home, even though they weren't party to the previous claims?
Not necessarily. One of my friends had his parked car hit by a drunk driver while he was not in it. His insurance premiums still went up after claiming for the damage.
- Not park their car in a garage overnight, exposing them to various damage to the car.
- Not park their car in a garage when they go downtown.
- Park their car near bars or other places statistically more likely to have drunk drivers around.
I'm not saying your friend did any of that and that it wasn't a complete freak accident, but on the scale of everyone that's being insured there's surely correlations like that that justify the premiums going up.
- not buy a house that's next door to a homeowner who rents the home to a person who is willing to drink and drive.
In case you didn't pick up on it, I'm calling your bullet points bullshit.
It's kind of like arguing over code, there's always going to be something you can nitpick as not optimal. There will always be something a person could possibly have done to prevent the issue (like committing suicide that morning), but it's a question of what's reasonable, and it is not unreasonable to expect a vehicle parked correctly in a parking spot to not be hit by another vehicle.
We're talking about insurance. There's certain reasonable activity that exposes you to heightened risk. Your car is at higher risk parked outside than in a garage.
Who should assume the cost of that heightened risk? Should the insurance company charge a person with a garage and a person without one the same premiums? Should there be no consequences for damage that could have been avoided by storing your car in a garage?
If the answer to those questions is "yes" you've created a moral hazard. Why would anyone store their car in a safer location if their insurance company will pay for any damages without their premiums going up?
To categorically say that this shouldn't be priced into insurance is to say that some buyers of insurance with higher risk profiles should be subsidized by those with lower risk profiles.
Now of course insurance companies are far from perfect. They have limited data to work with and it wouldn't be cost effective to figure out whether some individual customer had a one-off freak accident or not.
But over their entire customer base it's not an unreasonable approximation to connect your premiums to traffic incidents you were involved in, even though you were not at fault.
Not being at fault is not the same thing as not having exposed yourself to heightened risk. The police deals with the former, insurance companies deal with the latter.
By moving in next door to a drunk, how dare you.
The problem with your example, and the reason it doesn't apply to this particular discussion, is the rates will go up if a drunk drives into your garage door and hits your vehicle.
Is it because of a heightened risk, or is it because the insurance company removes you from their special "never had to pay anything out on you" list?
someone who gets a DUI is at an increased risk for wrecks, this is reasonable. Someone who is parked in a parking spot is not, that is not reasonable. The driver is already paying a rate based upon where they live, they have ALREADY paid for that risk.
What the heck? Can you elaborate on this? How is this even legal?
No, because you're paying for the peace of mind that comes with knowing that your worst-case scenario has been mitigated, on the off-chance it happens. Even if there is no disaster, you still benefited from not having to worry about it as much (knowing that the insurer would cover some portion of it).
Let's play a game: you're allowed to flip a coin ten times, and I will pay you $1 for each time it comes up heads[0]. Your expected value is $5, but you could make as much as $10 or as little as $0.
Someone else comes up to you and offers to pay you a guaranteed minimum of $2 at the end of the game. In exchange, you have to pay him $.25 every time you flip the coin, regardless of whether it comes up heads or tails.
In expectation, you're going to come out behind, but this game tightens the variance on your outcome. The worst case is that you lose $.50 (you pay the insurer ten quarters, flip ten tails, and receive $2 from the insurer). The best case is that you receive $7.50 (you pay the insurer ten quarters, flip ten heads, and receive $10 from me). The expected outcome is lower, but the range of possible outcomes is $8, not $10.
In the real world, you're insuring against money being taken from you, but mathematically it's the same, and I find it's easier to explain it this way. Most people have a more intuitive sense of earning money than losing it. Go figure - psychology is weird.
[0] One notable difference here is that the distribution of coin flips is approximately normal (binomial), whereas most insured events have a very long tail, with a mode of 0.
Now of course if they charge too high, then you could go to a competitor. But there is this CLUE database (I just learned about in a sibling comment here) which apparently is shared among insurers and depending on what is there (or what is not there! -- such as, it was really just an accident) you get labeled and dropped. So even a competitor might decide it is safe to stay away from you.
The irony is that the peace of mind I end up with when dealing with insurance companies of any sort is the dead certainty that they are going to screw me over just as hard as they possibly can, should I ever actually need the service they pretend to be in the business of selling, so I'd better give them as little as I can legally get away with and expect to handle the unlikely disaster on my own.
They don't expect profit from all their accounts, not a priori. However they choose to do business only with those account which they believe will be profitable. If They expect to loose money from an account, they'll choose not to do business with you.
We can debate if this is moral or not (some religions state it's not, i.e. the Amish as it is a form of gambling and discourages charity) but that's the game.
The purpose is clearly to discourage you from making a claim, which is what makes insurance a legal scam.
Which is why I'm skeptical of the AMA requirements to cover certain recurrent health services (annual checkups, contraception). If they're unavoidable than its not insurance!
Also, most health insurance plans have always covered checkups, contraception etc. The ACA specifies that these should be covered without copayments and deductibles. This is intended to be an incentive for the insured to get these kinds of preventative procedures instead of waiting till more costly problems develop. This tends to reduce the total cost of care to be reimbursed and (in theory) should lower premiums.
Which is an absurd way of thinking about it, but there you are.
This isn't a scam at all. That's exactly what you'd expect from insurance. Insurance isn't there to save you money. In fact, the expected value of an insurance plan is always going to be negative on an infinite time horizon[0].
The whole point of an insurance plan is to smooth risk: instead of the small possibility of a catastrophic failure, you take the relative certainty of having a fixed monthly payment (your premium). At a very high level, it's sort of similar to taking on consumer debt (credit card debt): in the long run, you'll always pay more than if you just paid cash up-front, because of the interest. The difference between credit cards and insurance is that you don't know when the expensive event is going to happen, so you agree to always pay the monthly payment regardless of what happens. And while banks charge interest based on the risk that you won't pay back and the opportunity cost of that money, insurers charge you a premium in exchange for the luxury of not having to guess whether next month will have an expensive event or not.
In your friend's case, the fact that she had two claims adjusted her relative risk profile, which means that her premium had to adjust accordingly.
[0] If it weren't, there'd be no point to going into business as an insurer, because you'd be guaranteed to make an operating loss. The only way you'd be able to make any money is based on the investments that you make, at which point you might as well eliminate the insurance side of your business and operate as a bank. (Interestingly, this is why the relationships between banks and insurers are so close in many countries, and also why most insurance markets - car insurance and health insurance[1] being two notable exceptions - have surprisingly slim profit margins).
[1] Health insurance is a special case because, strictly speaking, it's not really "insurance", but that's a whole separate discussion.