Thursday, April 21, 2011

Risk Profiles

RD brought up the discusion of risk profiles in a recent comment on my post about the difference between medical and auto insurance.

RD spoke quickly and was wrong. He claimed the difference between health and auto insurance was the risk profile. His claim was:

"the primary difference is the distribution within the dataset. Auto insurance costs having a fairly even distribution of costs through its dataset and a comparative marginal worst case cost cap, No insurance will pay to fix a car where the cost of fixing is higher then replacement."

The opposite is true. The distribution of claims with auto insurance is absolutely wild compared to health insurance. There is a large number of people who never have a claim against their auto insurance. A person driving a car through a crowd of pedestrians or causing a multicar pile up can rack up millions of dollars in liability in an instant.

Since everyone needs to spend money to maintain their health, there is less of a deviation in health expenses over one's life than with auto insurance.

There is often a great deal of variance per year throughout a person's life, however, the variance in medical expenses over a person's lifetime is not that great ... especially if we were to adopt preventive medicine.

The whole point of the Medical Savings and Loan is that, when you look at people from a whole life perspective, there is not as much variance in expenses as one would first suspect. When I did this experiment before, I discovered that there was a very nice bell curve with most people's life time health expenses sitting between reasonable figures like $300,000 to $600,000. With guidance, all but a few people could self fund their care.

This is not true of auto insurance where there is a wild difference in expenses. Each year only a fraction of people have claims. A small number of claims reach into the stratosphere.

The primary difference between health and auto liability insurance, as far as public policy is concerned, is that auto liability insurance is driven by the court system. Liability go through the court. If not for insurance, health care would be a matter of private contracts between patients and providers. Buying health insurance invites the court system and government into the care of one's health.

I know this for fact because, when I buy health care from my own pocket, I pay a bill negotiated with the doctor. When I use insurance I file a claim against a pool. Paying a bill is different from filing a legal claim. The first involves direct negotiation. The second is controlled by a court.

When I run a widow with an SUV full of orphans off the road, I am not involved in negotiation. I am involved in a court case. The amount paid gets decided by claims adjusters, appraisers and a judge with greater sympathy to widows and orphans than overweight computer dinks.

There is another substantial difference between health care and auto insurance. Auto insurance is about the liability of a single instance. Health insurance involves the whole series of things we do during our life to maintain our health.

I will write a second post on this matter in the near future.


RD said...

comparing the rate at which multi-car pile ups in the middle of large crowds(1 in 100,000-200,000) and cancer(1 in 3 to 7 depending on how you count) is like comparing the surface area of a dollar bill to that of a foot ball field and saying they are similar enough to be more or less the same.

Accidents like you describe in fact do happen they are simply statistically so uncommon that they don't amount to much, over 95% of auto claims are for light body damage and paint touch ups this is also the majority of the expenses auto insurance pays for. While Health issues are statistically common, but have a much wider mean variance and the claims that payout the most in a dollar amount are flipped from auto insurance as the top 5%-10% are where the most money is spend.

And having $300,000 to $600,000 dollars in an account at age 83 is a useless measure of the capacity of that model to handle health care. Health care issues are not distributed in such a way as to be kind enough to people to wait for their savings account to accrue enough interest over the years to be useful.

And I ran the numbers on your savings plan, you would be lucky to have $200,000-$250,000 in the account at death assuming you never used once through out your entirely life(unlikely).

The interest rates available in savings accounts isn't enough to be useful, and very few people can properly manage an investment portfolio to obtain rates that are more suited to grow such an savings pool.

And medical loans will never ever happen, the death rate of the people who need them make the entire concept absurd, to say nothing of the normal rate of bankruptcies, and likely higher non-repayment rate in such ridicules loans due to people with continuing care problems and lose of work due to disability.

y-intercept said...

Wikipedia claims that there were 33,808 deaths in 2009. I don't think that is statistically insignificant.

BTW, I don't think you ran any numbers against the formula I gave, because I have not given any formulas beyond the statement that I split the money into three piles: Savings Accounts, A Loan Reserve and Grants (The unused portion of loan researves falls into the savings accounts).

I did not give specific percentages because I would want a detailed actuarial analysis before setting such numbers. I do not have access to any numbers that I trust.

The basic idea is that people should have direct control over the first block of one's life time health care dollars. All my statement about numbers are tentative.

The actual number would be a portion of one's income ... using a progressive scale, of course. A person with a $10,000,000 life time earnings might be expected to foot the first four million dollars in health care expenses before receiving any sort of assistance, while a person with $500,000 might start receiving assistance after only $20,000 in medical bills.

RD's claim that some magickal property about the distribution of health expenses demands that the government be involved in every single transaction between doctor and patient is absolute nonsense.

The courts are involved in automobile accidents because such accidents automatically involve a dispute between the parties ... not because of a matheMAGICKal property of the distribution of claims.

If people had direct control over the first block of their lifetime medical expenses (about $250,000 per person) then we could get courts and government out of health care.