Comment by defen | Hacker News Reader

Comment by defen | Hacker News Reader

Parent Full thread

Insurance works best for rare, randomly distributed events that have a high cost. Fundamentally, for insurance to work as a business, it has to cost more than the expected payout. The reason it's still worthwhile is that money has non-linear utility. To use a made-up example from the car insurance world (numbers are made-up but I hope it conveys the general idea): a guaranteed "loss" (cost of premium) of $100/month over the course of 30 years (total cost: $36,000) is far preferable to having a 1% chance of losing $500,000 over that same time period if you seriously injure yourself or someone else in an accident (expected value: -$5,000). So, the average person is paying $36,000 to avoid an expected loss of $5,000, which doesn't seem to make sense in a naive analysis. But the issue is that the average person can plan/budget for a $100/month expenditure, whereas a $500,000 loss could mean that they lose their house and life savings and more, which is far too high of a risk for most people.

Things that insurance doesn't work well for: common events (e.g. routine medical care); non-randomly distributed events (that's why e.g. earthquake insurance in San Francisco is extremely expensive and has very high deductibles - if the Big One™ happens, it's going to affect everyone at the same time); inexpensive events (e.g. why insure a phone?)

> Things that insurance doesn't work well for: common events (e.g. routine medical care); non-randomly distributed events (that's why e.g. earthquake insurance in San Francisco is extremely expensive and has very high deductibles - if the Big One™ happens, it's going to affect everyone at the same time); inexpensive events (e.g. why insure a phone?)

I agree with this, but I want to clarify your language: insurance works very well for common events - what it doesn't work well for are predictable events (which is why insurance is a terrible model to apply to routine medical care).

The entire value proposition of insurance is to reduce (but not eliminate) uncertainty. If an event is perfectly predictable on a regular cadence, there's no value to insuring it.

Events that are common but have high variance are potentially worth insuring, for the same reason that a person whose earnings are highly variable (contractor, or tip-based work) might choose to use a service that evens outs those payments, even though it lowers their total take-home earnings (by charging a fee).

> which is why insurance is a terrible model to apply to routine medical care

Yet, insurance is the only ethical model to apply to (routine) medical care. A society in which only the rich(est) can afford the medical care their specific situation requires and the rest is left to fend for themselves cannot be considered humane.

Is universal healthcare funded with taxes unethical? The system used in most developed countries has nothing to do with insurance.

> Is universal healthcare funded with taxes unethical? The system in most developed countries has nothing to do with insurance.

Well, "universal healthcare funded with taxes" isn't strictly speaking the system in most developed countries either. Most countries have a mixture of private and public models.

The private part is optional and not widely used in many countries. And in some places it’s even illegal. The question stands: is the tax funded part unethical?

> The private part is optional and not widely used in many countries. And in some places it’s even illegal.

Going in descending order of population for EU countries:

* Germany: Mixture of private and public. Taxpayer funds cover about 75% of expenses. Private insurance can be used to cover the remainder.

* France: Mixture of private and public. Taxpayer funds only cover about 70% of expenses. Private insurance can be used to the remainder.

* UK: Arguably the most nationalized system, but still a heterogenous pool. NHS England ostensibly covers a little more than 90% of expenses, though it's complicated because the NHS sometimes subcontracts out to private management. NHS also has some of the worst ratings for specialized care (not routine care) among OECD countries, which is what other countries typically lean on private payments for.

* Italy: Mixture of private and public. The SSN doesn't cover 100% of expenses, and private insurance can be used to cover the remainder (or to pay for treatment in private practices). Similar to Medicare in the US, the SSN drives up reimbursements for private practices by setting minimum prices for private practices.

* Spain: Mixture of private and public. SNS doesn't cover all expenses, such as ambulances, dentists, pharmaceuticals, etc.

I could go on, but you get the idea. It's not really accurate to say that "the system in most developed countries has nothing to do with insurance", because most OECD countries require private payments for a non-trivial portion of care, and insurance is typically used to manage that portion of the costs.

Ok, let me be more precise. The universal healthcare system which pays for the large majority of healthcare costs in most developed countries does not work as an insurance (it’s paid by taxes or compulsory affiliation unrelated to actuarial risk). In some cases, one has the option to contract private insurance to pay for the parts not covered by the public system. It’s debatable to what extent this is comparable to actual healthcare insurance, because it’s not insuring the catastrophic events (that are already covered) but the relatively small and frequent out-of-pocket payments and some particular things like dental and vision care.

> Ok, let me be more precise. The universal healthcare system which pays for the large majority of healthcare costs in most developed countries does not work as an insurance (it’s paid by taxes or compulsory affiliation unrelated to actuarial risk)

As pointed out at the beginning of this thread, even in the US, health insurance is "insurance" in name only. Premiums are not set by actuarial risk (with the binary exception of smoking status).

That was the whole point of this subthread: there is no such thing as "actual healthcare insurance", because actual insurance is not a model that's applied to healthcare in any OECD country.

I don’t disagree. Healthcare insurance in the US is a fiction (especially now that catastrophic-only coverage is not an option). But other countries don’t even have this fiction of lives being saved by healthcare “insurance” because the one saving your life is the NHS, etc.

When mschuster91 said that insurance is the only ethical model it was not clear to me if he meant “insurance as in the US which is insurance in name only” or insurance as in “the model used in other countries which is not even called insurance”, hence me question.

It's a simple task to model a tax-funded healthcare system as an insured system (with monopsony), where the government is the insuror and the taxes that are directed to the are the premium. (And since the government is the sole payor in the system, it can unilaterally set the prices).

> It's a simple task to model a tax-funded healthcare system as an insured system (with monopsony), where the government is the insuror and the taxes that are directed to the are the premium

Unless you're proposing that taxes are based on risk levels, rather than realized income, then what. you are proposing is neither a simple task nor a true insurance model.

If that's what you're proposing, then it's an interesting thought experiment, but not one that applies to anything currently practiced anywhere in the world - and it's also something that would raise a lot of ethical objections.

Insurance doesn’t work well for common events if the administrative overhead is large compared to the economic loss insured.

Health insurance is both insurance and a discount/negotiation program. It'd be nice to disentangle the two.

Reply on news.ycombinator.com